Victor Michelon Alves EFEITO DO USO DO SOLO NA DIVERSIDADE

Victor Michelon Alves

EFEITO DO USO DO SOLO NA DIVERSIDADE E NA MORFOMETRIA DE BESOUROS ESCARABEÍNEOS

Tese submetida ao Programa de Pós- Graduação em Ecologia da Universidade Federal de Santa Catarina para a obtenção do Grau de Doutor em Ecologia. Orientadora: Prof.a Dr.a Malva Isabel Medina Hernández

Florianópolis 2018

AGRADECIMENTOS

À professora Malva Isabel Medina Hernández pela orientação e por todo o auxílio na confecção desta tese. À Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) pela concessão da bolsa de estudos, ao Programa de Pós- graduação em Ecologia da Universidade Federal de Santa Catarina e a todos os docentes por terem contribuído em minha formação científica e acadêmica. Ao professor Paulo Emilio Lovato (CCA/UFSC) pela coordenação do projeto “Fortalecimento das condições de produção e oferta de sementes de milho para a produção orgânica e agroecológica do Sul do Brasil” (CNPq chamada 048/13), o qual financiou meu trabalho de campo. Agradeço imensamente à cooperativa Oestebio e a todos os produtores que permitiram meu trabalho, especialmente aos que me ajudaram em campo: Anderson Munarini, Gleico Mittmann, Maicon Reginatto, Moisés Bacega, Marcelo Agudelo e Maristela Carpintero. Ao professor Jorge Miguel Lobo pela amizade e orientação durante o estágio sanduíche. Ao Museu de Ciências Naturais de Madrid por ter fornecido a infraestrutura necessária para a realização do mesmo. Agradeço também a Eva Cuesta pelo companheirismo e pelas discussões sobre as análises espectrofotométricas. À Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) pela concessão da bolsa de estudos no exterior através do projeto PVE: “Efeito comparado do clima e das mudanças no uso do solo na distribuição espacial de um grupo de insetos indicadores (Coleoptera: Scarabaeinae) na Mata Atlântica” (88881.068089/2014-01). Ao professor Fernando Vaz de Mello (UFMT) pela identificação dos escarabeíneos. Aos integrantes do Laboratório de Ecologia Terrestre Animal (LECOTA) especialmente Pedro da Silva e Renata Campos pelas discussões e críticas ao meu trabalho. Ao Eduardo Giehl pela ajuda nas análises e no campo. Aos membros da pré-banca e da banca examinadora por terem se dedicado a avaliar meu trabalho. A minha mãe Salete, meu pai Luis e minha namorada Liziane, por sempre estarem comigo me apoiando.

RESUMO

A heterogeneidade do hábitat está associada a uma maior oferta de recursos e condições, o que permite a expansão do nicho das espécies, promovendo a coexistência e acarretando em comunidades mais ricas e em uma maior complexidade ecossistêmica e ecológica. Quando esse padrão heterogêneo é alterado, e.g. conversão em monocultivos, ocorre uma simplificação do ecossistema e as espécies com nichos mais estreitos são localmente extintas. A Mata Atlântica do sul do Brasil (Floresta Ombrófila Mista) é caracterizada por essa heterogeneidade e é considerada um dos principais hotspots de diversidade e, devido principalmente à expansão agrícola, resta atualmente menos de 10% de sua cobertura original. Além de alterar drasticamente o ambiente, podem ser utilizados cultivos de organismos geneticamente modificados (OGM) e pouco se sabe sobre o efeito destes OGMs nas comunidades. Os besouros escarabeíneos (Coleoptera: Scarabaeinae) são amplamente utilizados como indicadores ecológicos por responderem rapidamente às mudanças provocadas pelo uso da terra. Além disso, também desempenham importantes funções ecossistêmicas (e.g. ciclagem de nutrientes) devido ao hábito alimentar copronecrófago e ao comportamento de enterrar o alimento no solo. A tese está dividida em quatro capítulos com amostragens realizadas na região de São Miguel do Oeste (26°43'31''S, 53°31'05''O), estado de Santa Catarina, sul do Brasil, no verão de 2015. O primeiro capítulo relata os possíveis impactos de três tipos de variedades de milho - convencional, crioulo e transgênico - sobre as assembleias de escarabeíneos que habitam os próprios cultivos e também as florestas adjacentes. No total, foram coletados 2126 indivíduos de 38 espécies, e foram encontradas diferenças na estruturação das assembleias que foram afetadas pelos diferentes tipos de milho. As comunidades das matas adjacentes aos cultivos transgênicos apresentaram menor riqueza em relação às matas adjacentes aos cultivos crioulos. Isso mostra que a intensificação do uso de transgênicos pode afetar negativamente as comunidades, levando à perda de espécies, e consequentemente à perda das funções ecossistêmicas exercidas pelos escarabeíneos. No segundo capítulo trabalhamos com a hipótese de que o milho transgênico pode afetar a aptidão (fitness) das espécies. Assim, escolhemos a espécie Canthon quinquemaculatus por apresentar um amplo nicho, refletindo em sua ampla distribuição espacial, com indivíduos tanto nas áreas florestais quanto nos cultivos de milho. Seis populações foram amostradas: três de cada tipo de milho e três de cada mata adjacente e a forma do corpo foi

extraída com a utilização de métodos morfométricos através do uso de 15 marcos anatômicos. Os indivíduos que habitam os cultivos transgênicos apresentaram diferenças morfométricas em relação aos da mata adjacente. Indivíduos do milho transgênico são levemente mais ovais e possuem uma retração na região abdominal. Esta modificação morfométrica pode ser um possível efeito subletal do milho transgênico nessa espécie, podendo ter implicações para o fitness da mesma, interferindo em sua sobrevivência, reprodução, e nos serviços ecossistêmicos de remoção de matéria orgânica. No terceiro capítulo propusemos observar as comunidades através das lentes da diversidade morfométrica, com cada espécie ocupando um determinado espaço de forma dentro de um universo morfométrico. Consideramos a diversidade morfométrica como o tamanho do hipervolume morfométrico ocupado por uma comunidade. Além de quantificar o hipervolume, também calculamos os buracos remanescentes nesse hipervolume, que são regiões vazias e que, em termos matemáticos, poderiam estar ocupadas. Observamos que a diversidade morfométrica está positivamente correlacionada com a diversidade taxonômica e negativamente com o tamanho dos buracos. As comunidades das áreas florestais apresentaram maiores valores de diversidade morfométrica e menores buracos em relação às comunidades dos cultivos de milho. Assim, o aumento dos buracos nas comunidades dos cultivos indica a perda de espaços de forma nessas áreas, que é um reflexo da extinção local de espécies. No quarto capítulo trabalhamos com a hipótese de que distintas espécies de Canthon teriam respostas espectrofotométricas diferentes dependo do habitat que se encontram ao longo da paisagem. Selecionamos duas espécies que habitam preferencialmente as áreas agrícolas (cultivos de milho): C. chalybaeus e C. tetraodon, duas de mata nativa: C. angularis e C. lividus lividus e uma espécie que habita ambas as áreas: C. quinquemaculatus. Todas as espécies apresentaram um padrão semelhante: as radiações de maiores comprimentos de onda e menores frequências são transmitidas para dentro do corpo, já as radiações de menores comprimentos de onda e maiores frequências (e.g. ultravioleta) são absorvidas. No entanto, a espécie C. quinquemaculatus apresentou maiores taxas de transmitância na radiação visível e infravermelha, e essa maior permeabilidade térmica de C. quinquemaculatus pode estar associada com o amplo nicho da espécie, estabelecendo populações tanto nas áreas abertas quanto nas áreas florestais.

Palavras-chave: Escarabeíneos. Espectrofotometria. Hipervolume. Morfometria Geométrica. Nicho Ecológico. Transgênicos. Zea mays.

ABSTRACT

The habitat heterogeneity provides more resources, allowing the species niche expansion, promoting coexistence and culminating in richer communities and greater ecosystemic and ecological complexity. When this heterogeneous pattern is modified, e.g. conversion to monocultures, ecosystem simplification occurs and species with narrower niches are locally extinct. The Atlantic Forest of southern Brazil is characterized by this heterogeneity and is considered one of the main diversity hotspots and, due mainly to agricultural expansion, there is currently less than 10% of its original cover. In addition to altering the environment, crops of genetically modified organisms (GMO) may be used, and little is known about the effect of these GMOs on communities. Dung beetles (Coleoptera: Scarabaeinae) are widely used as ecological indicators because they respond to land use changes, and also are importants in ecosystem functions (e.g. nutrient cycling) due its behavior of removing and burying organic matter. The thesis was divided into four chapters made in the region of São Miguel do Oeste (26°43'31''S, 53°31'05''W), Santa Catarina state, southern Brazil, in the summer of 2015. The first chapter reports the potential impact of three types of maize varieties - conventional, creole and transgenic - on the dung beetles assemblages that inhabit the crops and also in adjacent forests. A total of 2126 individuals were collected from 38 species, and differences were found in the assemblages that were affected by different types of maize. The forest communities adjacent to the transgenic crops presented lower richness than the forests adjacent to the creole crops. This shows that the transgenic maize use may be negatively affect the communities, leading to diversity loss and therefore the ecosystemic functions provided by dung beetles may have been lost. In the second chapter we work with the hypothesis that transgenic maize can affect the species fitness. Thus, we chose the species Canthon quinquemaculatus because it has a broad niche, reflecting in its wide spatial distribution, with individuals both in the forest areas and in the corn crops. Six populations were sampled: three maize crop populations each under a different seed type, and three populations of adjacent forests. The individuals that inhabit the transgenic crops presented morphometric differences in relation to the adjacent forest. Insects in transgenic maize are more oval and have a retraction in the abdominal region, this result shows the possible effect of transgenic crops on non-target species. This may have implications

for the ecosystem service of organic matter removal, carried out by these organisms. In the third chapter we proposed to evaluate the communities through the lenses of morphometric diversity, with each species occupying a body shape space within a morphometric universe. We consider morphometric diversity as the size of the morphometric hypervolume occupied by a community, in addition to quantifying hypervolume, we also calculate the remaining holes in this hypervolume, which are empty areas with no species. We found that the taxonomic diversity positively correlates with the morphometric diversity, and negatively correlates with the size of the holes. We verified that forest communities have higher values of morphometric diversity and smaller holes in the hypervolume than the maize cultivation communities, demonstrating that species extinction reduces community body shape spaces. In the fourth chapter we work with the hypothesis that different Canthon species would have different spectrophotometric patterns depending on their habitat. We selected two species that preferentially inhabit forest areas (Canthon angularis and Canthon lividus lividus), two species preferentially inhabit open areas (Canthon chalybaeus and Canthon tetraodon) including agricultural crops, and one species does not present a clear habitat preference and can be found in both habitats (Canthon quinquemaculatus). All the species show a similar pattern in which the light from shorter wavelengths and higher frequencies is almost entirely absorbed by the elytra, while radiation from longer wavelengths and lower frequencies can mostly pass through the elytra. However, C. quinquemaculatus seems to have significantly higher rates of reflectance and transmittance in the visible and near-infrared spectrum. This different pattern found in C. quinquemaculatus may be associated with its capacity to establish populations both in agricultural and forest areas.

Keywords: Dung beetles. Ecological Niche. Geometric Morphometric. Hypervolume. Spectrophotometry. Transgenic. Zea mays.

LISTA DE FIGURAS

INTRODUÇÃO GERAL Figura 1 - Interpretação do nicho Hutchinsoniano definido por três dimensões, a esfera (nicho) representa o quanto de cada fator é necessário para determinada espécie sobreviver (Adaptado de Chase & Leibold 2001)...... 22

CAPÍTULO I

Figure 1 - Location of dung beetle samples in the region of São Miguel do Oeste, state of Santa Catarina, Brazil. Each site (points magnified in the figure) represents a block of samples consisting of two landscape units: maize fields and adjacent forest fragments. Blocks were replicated five times for each type of maize grown in the landscape: conventional, creole and transgenic...... 41

Figure 2 - Non-metric multidimensional scaling (NMDS) ordination plot for dung beetle assemblages sampled in 30 landscape units (15 maize fields and 15 nearby forest fragments) and where three types of maize are grown in the region São Miguel do Oeste, state of Santa Catarina, Brazil...... 49

Figure 3 - Dissimilarity of dung beetle assemblages occurring in forest fragments nearby maize fields where three types of maize are grown (conventional, creole and transgenic) in the region of São Miguel do Oeste, state of Santa Catarina, Brazil. Species drawn are the most representative in each landscape unit and account for ~ 60% of the differences between them...... 50

Figure 4 - Dissimilarity of dung beetle assemblages occurring in fields where three types of maize are grown (conventional, creole and transgenic) in the region of São Miguel do Oeste, state of Santa Catarina, Brazil. Species drawn are the most representative in each landscape unit and account for ~ 60% of the differences between them...... 51

Figure S1 - Landscapes of the blocks with two landscape units (cultivation and forest remnants) of conventional corn areas in the region of São Miguel do Oeste, state of Santa Catarina, Brazil...... 65

Figure S2 - Landscapes of the blocks with two landscape units (cultivation and forest remnants) of creole corn areas in the region of São Miguel do Oeste, state of Santa Catarina, Brazil...... 66

Figure S3 - Landscapes of the blocks with two landscape units (cultivation and forest remnants) of transgenic corn areas in the region of São Miguel do Oeste, state of Santa Catarina, Brazil...... 67

Figure S4 - Scheme of blocks consisting of two landscape units: maize fields and nearby forest fragments. Blocks were replicated five times for each type of maize grown in the landscape (conventional, creole and transgenic), totaling 15 blocks and 30 landscape units where dung beetles were sampled in the region of São Miguel do Oeste, state of Santa Catarina, Brazil...... 68

Figure S5 - Representation of the pitfall traps arrangement to dung beetle collection in the region of São Miguel do Oeste, SC, Brasil. Each area had 10 traps: five baited with human feces and five with pork carrion. The traps with same bait were placed 10 meters from each other and five meters with different baits. The sampling unit was the pair of traps...... 69

Figure S6 - Residuals resulting from the fitted multivariate generalized linear model (GLMmv)...... 70

CAPÍTULO II

Figure 1 - Landmarks (red dots) used in shape analysis of C. quinquemaculatus. (A) Lateral view; (B) Graphical representation of body shape based on 15 landmarks, adapted from Hernández et al. [8]. Landmarks in parentheses correspond to the same region on the other side of the body...... 77

Figure 2 - Position of C. quinquemaculatus populations in morphometric space formed by two main axes of Canonical Variable Analysis, collected in São Miguel do Oeste, SC, Brazil. Conventional (Co), Creole (Cr), and Transgenic (Tr); ellipses indicate a 95% confidence interval...... 79

Figure 3 - (A) 2D Graphical reconstruction of C. quinquemaculatus body shape. Gray lines show body shape of transgenic corn crop

populations, and black lines show adjacent forest populations. (B) 3D Graphical reconstruction of C. quinquemaculatus body shape. Green dots: transgenic corn populations. Pink dots: adjacent forest populations...... 81

Figure A1 - Experimental design scheme used for the captured individuals of C. quinquemaculatus in maize crops and adjacent forests of Atlantic Forest in the region of São Miguel do Oeste, SC, Brazil.....90

CAPÍTULO III

Figure 1 - Landmarks (red points) used in the construction of the body shape. (A) Dorsal view; (B) Ventral view; (C) Lateral view; (D) Graphical representation of body shape based on the 15 landmarks in three dimensions, adapted from Hernández et al. (2011)...... 97

Figure 2 - Morphometric universe and global hypervolume. (A) Global morphometric universe of dung beetles, with each of the 38 species occupying a specific region within the universe, some genera were drawn to represent the variation in body shape; (B) hypervolume of the global community (red points) composed by 38 species and hypervolume holes (blue dots)...... 100

Figure 3 - Morphometric niche represented in the first two axes of the Principal Component Analysis. Drawn are examples of morphometric niche and width of morphometric niche (95% C.I) of six species that have overlapping of their morphometric niches, however, are differentiated in other ecological niche variables……………….……101

Figure 4 - Morphometric diversity (MD) (red points) and holes (HS) (green points) of the 15 dung beetle communities from the forest fragments, collected in the region São Miguel do Oeste, SC, Brazil. The hypervolume volume was considered as the value of morphometric diversity...... 102

Figure 5 - Morphometric diversity (MD) (red points) and holes (HS) (green points) of the 15 dung beetle communities from maize cultivation areas, collected in the region São Miguel do Oeste, SC, Brazil. The hypervolume volume was considered as the value of morphometric diversity...... 102

Figure 6 - Relationship between morphometric diversity to species diversity (A); and relationship of hole size to species diversity (B) in 30 communities of dung beetles, 15 forest communities and 15 maize cultivation communities, collected in São Miguel do Oeste, SC, Brazil...... 104

CAPÍTULO IV

Figure 1 - Elytra color variation of the five dung beetle species used in the study: (A) C. angularis (B) C. lividus lividus (C) C. chalybaeus (D) C. tetraodon and (E) C. quinquemaculatus; and estimated dung beetle body measurements, the shaded region indicates the anatomical position of the elytra removed for analysis...... 125

Figure 2 - Absorbance, reflectance and transmittance percentages across the complete spectrum for the five considered dung beetle species: (A) C. angularis, (B) C. lividus lividus, (C) C. chalybaeus, (D) C. tetraodon and (E) C. quinquemaculatus. Shaded areas represent variation in the valued obtained for the different individuals...... 127

Figure 3 - Absorbance (squares), transmittance (circles), reflectance (triangles) and adjusted mean values (± 95% CI) of the different species in the (A) visible electromagnetic spectrum and in the (B) near infrared electromagnetic spectrum when all the considered predictors are held at their means (i.e. controlling for the effect of the three covariates included in the model; body weight, volume and elytra thickness).....130

Figure S1. Visible spectral reflectance of the elytra of the different studied species...... 144

LISTA DE TABELAS

CAPÍTULO I

Table 1 - Dung beetle species sampled in 30 landscape units (15 maize fields and 15 nearby forest fragments) and where three types of maize are grown (CO: Conventional; CR: Creole; TR: Transgenic) in the region of São Miguel do Oeste, state of Santa Catarina, Brazil...... 44

Table 2 - Abundance, observed Richness (with confidence intervals of 95%), estimated Richness Chao-1 and Sample coverage of dung beetle assemblages in forest fragments and cultivation areas with three different maize types: Conventional (CO), Creole (CR) and Transgenic (TR), collected in the region of São Miguel do Oeste, State of Santa Catarina, Brazil...... 47

Table 3 - Scores of the first two axes of the principal component analysis (PC1 and PC2) of the variables describing vegetation structure of the forest fragments nearby maize fields in the region of São Miguel do Oeste, state of Santa Catarina, Brazil. Tree functional groups area as follow: pioneer (P), secondary light-demanding (SL), and secondary shade tolerant tree species (SS)...... 52

Table 4 - Explanatory variables used in the generalized linear model for multivariate responses (GLMmv) to explain differences in composition and abundance patterns of dung beetle assemblages in the region of São Miguel do Oeste, state of Santa Catarina, Brazil...... 52

Table S1. Use of herbicides, insecticides and fungicides in 15 maize fields where three types of maize are grown (CO: Conventional; CR: Creole; TR: Transgenic) in the region of São Miguel do Oeste, state of Santa Catarina, Brazil...... 61

Table S2. Tree species found in 15 Atlantic Forest fragments nearby maize fields in the region of São Miguel do Oeste, state of Santa Catarina, Brazil. The number of trees of each species found in landscapes with distinct types of maize are indicated (CO: conventional, CR: creole and TR: transgenic maize). Tree functional groups area as follow: pioneer (P), secondary light-demanding (SL), and secondary shade tolerant tree species (SS)...... 62

CAPÍTULO II

Table 1 - Percentage of correct classification of C. quinquemaculatus populations based on body morphology and discriminant analysis (DA)...... 80

CAPÍTULO III

Table S1 - Values of morphometric diversity (hypervolume volume, HV) and hole size (HS) followed by confidence intervals (CI) and coefficients of variation (CV). Forest fragment communities (FF); maize crops communities (MC)...... 113

Table S2 - Dung beetle species collected in the region of São Miguel do Oeste, SC, Brazil. Body shape values in the first principal components axes (PC1, PC2, PC3)...... 115

Table S3 - ANCOVA results of the variables that affect the morphometric diversity (in ln) of 30 dung beetles communities, 15 forest communities and 15 maize cultivation communities, collected in São Miguel do Oeste, SC, Brazil...... 116

Table S4 - ANCOVA results of the variables that affect the morphometric holes size (in ln) of 30 dung beetles communities, 15 forest communities and 15 maize cultivation communities, collected in São Miguel do Oeste, SC, Brazil...... 117

CAPÍTULO IV

Table 1 - Mean values (± 95% confidence interval) of absorbance (a), transmittance (t) and reflectance (r) measurements under near infrared (NIF), visible (VIS) and ultraviolet (UV) radiation of the five considered species of dung beetles...... 128

Table 2 - GLM results of transmittance, reflectance and absorbance values in the visible (VIS) wavelength spectrum (391-749 nm)...... 129

Table 3 - GLM results of transmittance, reflectance and absorbance values in the near infrared (NIF) wavelength spectrum (750-1400 nm)……………………………………………………………………131

Table S1- Abundance values collected for the five considered species of dung beetles in the surveyed forest fragments (FF) and maize crops (MC) within an agricultural landscape located in the region of São Miguel do Oeste, SC, Brazil………………………………………….141

Table S2 - Mean values (± 95% confidence interval) of the considered morphological variables used as covariates for five dung beetles species……………………………………………………………...... 143

SUMÁRIO

INTRODUÇÃO GERAL ...... 21 NICHO ECOLÓGICO ...... 21 GRUPO FOCAL ...... 25 USO DO SOLO ...... 27 HIPÓTESES ...... 32 OBJETIVOS ...... 33 OBJETIVO GERAL ...... 33 OBJETIVOS ESPECÍFICOS ...... 33

ARTIGO I: DUNG BEETLES AND THE CONSERVATION OF DIVERSITY IN AN AGRICULTURAL LANDSCAPE WITH MAIZE FIELDS AND ATLANTIC FOREST REMNANTS: THE INFLUENCE OF CREOLE, CONVENTIONAL AND TRANSGENIC CROPS ...... 35

ARTIGO II: MORPHOMETRIC MODIFICATIONS IN CANTHON QUINQUEMACULATUS CASTELNAU 1840 (COLEOPTERA: SCARABAEINAE): SUBLETHAL EFFECTS OF TRANSGENIC MAIZE? ...... 71

ARTIGO III: LOCAL EXTINCTIONS MAY BE EVIDENCED BY THE HOLES OF THE MORPHOMETRIC HYPERVOLUME IN DUNG BEETLES COMMUNITIES ...... 91

ARTIGO IV: ELYTRA ABSORB ULTRAVIOLET RADIATION BUT TRANSMIT INFRARED RADIATION IN NEOTROPICAL CANTHON SPECIES (COLEOPTERA, SCARABAEINAE) ...... 119

CONCLUSÕES GERAIS ...... 145

REFERÊNCIAS ...... 147

INTRODUÇÃO GERAL

NICHO ECOLÓGICO

O nicho é um dos pilares da teoria ecológica (ROOT, 1967), porém uma série de definições imprecisas (e.g. nicho = “profissão da espécie”) e diferentes interpretações ao longo do tempo dificultaram a aplicabilidade desse conceito. Diferentes literaturas, mesmo as mais tradicionais dentro da ecologia, trazem diferentes conceitos de nicho (ver RICKLEFS 1996; ODUM 1997). Essa dificuldade em conceituar o nicho é o reflexo de seu histórico, já que, ao longo de 40 anos, diferentes ecólogos contribuíram com visões distintas para a construção deste conceito. É amplamente aceito que Joseph Grinnell em 1917 fundou o conceito de nicho (CHASE & LEIBOLD, 2001) em seu trabalho The niche-relationships of the California thrasher. O conceito proposto por Grinnell (Nicho Grinnelliano), também conhecido como nicho ambiental, é determinado pelo hábitat no qual a espécie vive, portanto o nicho é moldado pelas características ambientais no qual a espécie se encontra (SOBERÓN, 2007). O Nicho Grinnelliano é ocupado por apenas uma espécie e normalmente está associado a amplas escalas geográficas (GRINNELL, 1917; SOBERÓN, 2007). Como a abordagem Grinnelliana considera o nicho como uma propriedade do ambiente (e não da espécie), cunhou-se o conceito de “nicho vago” que são regiões geográficas onde não existem espécies adaptadas, porém esse conceito já não é mais compatível com a teoria de nicho mais moderna (CHASE & LEIBOLD, 2001). Dez anos após o conceito Grinnelliano, Charles Elton, em 1927, desenvolveu um novo conceito de nicho. O Nicho Eltoniano foca no efeito da espécie no ambiente, e não no efeito do ambiente na espécie (que era o caso do nicho Grinnelliano). Desta forma, o Nicho Eltoniano se relaciona com os atributos funcionais de determinada espécie, com sua posição na cadeia alimentar e com seu impacto no ambiente. O conceito Eltoniano está associado ao uso de recursos, com base nas interações ecológicas, por isso relacionado ao modo de vida e função (CHASE & LEIBOLD, 2001). Posteriormente, em 1958, George Evelyn Hutchinson revolucionou o conceito de nicho, o tornando quantificável através de métodos matemáticos mais robustos. O nicho Hutchinsoniano é a soma de todos os fatores ambientais que atuam sobre uma determinada espécie. Portanto, o nicho passa a ser um hipervolume n- 22

dimensional ocupado por uma espécie dentro de determinado espaço n- dimensional (HUTCHINSON 1957, Figura 1).

Figura 1 - Interpretação do nicho Hutchinsoniano definido por três dimensões, a esfera (nicho) representa o quanto de cada fator é necessário para determinada espécie sobreviver (Adaptado de Chase & Leibold 2001).

O nicho é propriedade da espécie e o hipervolume n-dimensional ocupado pela espécie reflete as suas necessidades para a sobrevivência; as dimensões do nicho podem ser diferentes condições bióticas, como densidade de predadores e/ou recursos, ou abióticas, como temperatura, radiação solar, entre outros (BLONDER et al., 2014). Hutchinson trouxe um maior dinamismo para o conceito do nicho, pois tanto o nicho Grinnelliano quanto o Eltoniano eram estáticos, já Hutchinson propôs um nicho dinâmico, que pode aumentar ou diminuir, ao longo do tempo evolutivo. Hutchinson também diferenciou o nicho fundamental (NF) do nicho realizado (NR). O NF engloba todas as condições e recursos que permitem uma determinada espécie sobrevier e reproduzir, excluída a existência de predadores e competidores. Por sua vez, o NR corresponde a um intervalo mais limitado dentro do NF, quando é considerada a presença de competidores e predadores (SOBERÓN & PETERSON, 2005; SOBERÓN & NAKAMURA, 2009). Porém quando são consideradas interações positivas (e.g. mutualismo) o NR pode se expandir e ser maior do que o NF (PULLIAM, 2000). Tanto o NF quanto o NR são debatidos até hoje dentro da teoria ecológica, e recentemente foi proposto o “nicho existente” (NE) que seria a intersecção do NR com as condições bióticas e abióticas em determinado tempo “t”, ou seja, o NE considera determinado intervalo temporal da existência da espécie (SOBERÓN & ARROYO-PEÑA, 2017). Muitos métodos matemáticos já foram propostos para mensurar o hipervolume; desde os mais simples como a Análise de Componentes 23

Principais (ACP) (RICKLEFS & TRAVIS, 1980) até os mais complexos como o Convex hull (CORNWELL et al., 2006). Mas esses métodos violam algumas premissas do hipervolume e podem não representá-lo fielmente, pois a ACP assume que o hipervolume possui distribuição multinormal enquanto que o Convex hull é muito sensível à presença de outliers (BLONDER et al., 2014). Outra questão importante é que os métodos supracitados não detectam os buracos do hipervolume, resultando em hipervolumes inflacionados. Os buracos são regiões vazias do hipervolume, representando os desvios das expectativas, e a detecção dos buracos permite o levantamento de hipóteses sobre processos ecológicos e evolutivos que estejam ocorrendo fora do esperado (BLONDER, 2016). Atualmente, um método que vem sendo utilizado na detecção dos buracos é a estimativa de densidade de Kernel (EDK), que mensura o hipervolume através de uma probabilidade de distribuição de pontos, criando um Expectation Convex limitado por uma largura de banda (BLONDER et al., 2014). Deste modo os buracos são calculados pela diferença entre o hipervolume observado e o esperado ou Expectation Convex (BLONDER et al., 2014; BLONDER, 2016). Além da EDK ser utilizada em nível populacional, a mesma pode ser utilizada em nível de comunidade, sendo possível calcular o hipervolume total ocupado por determinada comunidade e também os buracos remanescentes no hipervolume. A competição é uma das interações ecológicas que estruturam as comunidades (TILMAN, 1982; TILMAN, 2004) e, assim como a coexistência, está intimamente ligada ao nicho (GAUSE, 1936; MACARTHUR & LEVINS, 1964), pois a competição e a coexistência são reguladas pelo grau de sobreposição do nicho das espécies (MACARTHUR & LEVINS, 1967; MAY, 1973; PIANKA, 1973). Espécies com o mesmo nicho (ou com alta sobreposição das variáveis do nicho) irão competir se os recursos forem escassos, assim a partilha de recursos permite a coexistência entre as espécies que difiram em seus nichos ecológicos (CONNELL, 1980), ao menos em uma dimensão (HUTCHINSON, 1957). Espécies coexistentes podem se diferenciar em várias dimensões do nicho, como o horário de atividade (HERNÁNDEZ, 2002), especialização por recurso (BEDOUSSAC et al., 2007) e preferência de micro-habitat (DOUGLAS & MATTHEWS, 1992). A partilha de recursos é um dos principais determinantes da diversidade de espécies coexistentes (PIANKA, 1974) e pode ser evidenciada na própria especialização pelo recurso (BEDOUSSAC et al., 2007) ou em diferentes métodos de exploração dos recursos (MACARTHUR, 1958), 24

levando à diferenciação de nichos e à consequente diminuição da competição (MACARTHUR & LEVINS, 1967). A partilha de recursos pode ser evidenciada externamente pela forma ou tamanho do corpo das espécies (SPENCER, 1995; HERNÁNDEZ et al., 2011). Portanto, a morfologia pode caracterizar o nicho da espécie, já que diferenças morfológicas estão associadas a diferentes variáveis do nicho (tanto Grinnelliano quanto Eltoniano), como a alimentação (INOUYE, 1980), predação (WEST et al., 1991), uso do hábitat (GASTON et al., 2012) e forrageio (MICHAUX et al., 2007). Portanto, o uso de métodos robustos para a mensuração morfológica (e.g. morfometria geométrica) pode ser muito útil na descrição do nicho das espécies e consequentemente uma importante ferramenta para avaliar as relações do nicho com a diversidade e a aptidão das espécies (fitness). Além disso, o nicho pode servir como uma base para avaliar as relações entre as espécies de uma comunidade: as comunidades podem ser vistas como um espaço de nicho total que pode ser preenchido e as espécies adequam seus nichos a esse espaço, de modo que comunidades mais ricas tendem a ter espécies com um maior estreitamento de nicho (COHEN, 1977). Esse estreitamento de nicho (ou especialização) pode ser evidenciado pela forma do corpo (KEAST & WEBB, 1985), resultando em uma relação entre a diversidade taxonômica e a diversidade morfológica. Desse modo, a representação do morfoespaço através da união das métricas do hipervolume (calculadas com a EDK) e a morfometria geométrica pode ser uma abordagem aplicável no estudo das relações entre o nicho e a diversidade. A aptidão (ou fitness) de um indivíduo depende da relação entre seu fenótipo e o ambiente (LEVINS, 1983). O fitness pode ser “mensurado” pela morfologia (tamanho ou forma do corpo), já que alterações morfológicas podem afetar drasticamente a reprodução (FORSMAN & SHINE, 1995; ROWE & ARNQVIST, 2011), deslocamento (PETERS, 1983) e o desenvolvimento (WINANS & NISHIOKA, 1987). Portanto, a relação nicho-aptidão também pode ser abordada através das técnicas de morfometria geométrica, pois estressores ambientais podem alterar a morfologia dos indivíduos de determinada população, afetando seu fitness e consequentemente podendo modificar a amplitude do seu nicho. As comunidades são compostas por uma parcela menor de espécies do que o pool de espécies potencialmente colonizadoras, isso devido aos diferentes filtros bióticos a abióticos que atuam sobre as espécies (VELLEND, 2010). Espécies com maior amplitude de nicho possuem maior taxa de dispersão (MCCAULEY et al., 2008) e menor 25

risco de serem localmente extintas (MCCAULEY et al., 2014). A capacidade de uma espécie de sobrepujar os diferentes filtros está diretamente associada ao seu nicho: espécies com nicho mais amplo podem suportar diferentes condições ambientais, já um nicho estreito levaria à extinção local frente a uma perturbação ambiental irreversível. Por exemplo, quando uma área de floresta é convertida em monocultivo aumenta-se a incidência da radiação solar, eliminando localmente as espécies que não conseguem suportar essa maior radiação, portanto, a radiação torna-se um importante filtro ambiental e espécies de nicho mais amplo, que conseguem habitar tanto as áreas de cultivo quanto as florestais, provavelmente devem possuir características fotobiológicas diferentes das espécies com nicho mais estreito. Se esse monocultivo é de um organismo geneticamente modificado (OGM) podemos ter um segundo filtro, que algumas espécies podem não atravessar, resultando em comunidades mais pobres. Portanto, uma abordagem baseada no nicho ecológico pode nos ajudar a entender tanto o efeito da conversão agrícola como o efeito específico do milho transgênico sobre as comunidades.

GRUPO FOCAL Os coleópteros constituem a maior ordem de insetos, com cerca de 350.000 espécies descritas (BOUCHARD et al., 2011) correspondendo a 40% das espécies descritas do subfilo Hexapoda (TRIPLEHORN & JONHSON, 2011). No Brasil já foram registradas cerca de 30 mil espécies, distribuídas em 105 famílias (CASARI & IDE, 2012). Dentre as famílias de Coleoptera, destaca-se Scarabaeidae com cerca de 30 mil espécies descritas, estando atualmente dividida em 13 subfamílias, oito com registros para o Brasil. A subfamília Scarabaeinae possui mais de 6,2 mil espécies descritas (TARASOV & GÉNIER, 2015), com 726 espécies no Brasil (VAZ-DE-MELLO, 2018) e, somente em áreas de Mata Atlântica no estado de Santa Catarina, há registo de 98 espécies (HERNÁNDEZ et al., submetido). Muitas espécies de escarabeíneos possuem nichos estreitos, resultando em alta especificidade (HALFFTER, 1991), havendo muitas espécies de ecossistemas florestais que não conseguem estabelecer populações em áreas abertas (KLEIN, 1989). Desse modo, as comunidades apresentam padrões distintos de organização quando presentes em fragmentos conservados ou em áreas deterioradas pela ação humana (HALFFTER & FAVILA, 1993; GARDNER et al., 2008a), respondendo a mudanças provocadas no ambiente pelo uso da terra (HALFTTER & FAVILA, 1993; revisão em NICHOLS et al., 26

2007). Além disso, os escarabeíneos estão intimamente relacionados em processos chave dos ecossistemas, sendo normalmente mais afetados que outros organismos (NICHOLS et al., 2007). Devido a essa sensibilidade às alterações no hábitat, os escarabeíneos são utilizados como um grupo indicador (HALFFTER & FAVILA, 1993). Outros fatores como a fácil captura e o baixo custo associado à amostragem, contribuem para que os escarabeíneos sejam amplamente utilizados como indicadores ecológicos (GARDNER et al., 2008b). Os escarabeíneos possuem hábito alimentar copronecrófago, detritívoro, se alimentando principalmente de carcaças e fezes de vertebrados (HALFFTER & MATTHEWS, 1966; HANSKI & CAMBEFORT, 1991). Deste modo, desempenham importantes funções ecossistêmicas, principalmente na ciclagem de nutrientes (HALFFTER & MATTHEWS, 1966; revisão em NICHOLS et al., 2008). Algumas espécies contribuem na dispersão de sementes, constituindo um papel importante na regeneração florestal (ESTRADA & COATES- ESTRADA, 1991). Os escarabeíneos também são importantes na regulação das propriedades químicas e físicas do solo, ao permitir a aeração e a hidratação do mesmo devido ao hábito de construir túneis (HALFFTER & MATTHEWS, 1966; SIMMONS & RIDSDILL- SMITH, 2011). Em áreas agrícolas, 80% do nitrogênio presente no esterco é perdido por volatilização, mas quando há escarabeíneos em quantidade suficiente para alocar esse material no solo a perda é reduzida entre 5 e 15% (GILLARD, 1967). Ao repor nutrientes no solo, não só o nitrogênio, mas também fósforo, cálcio e potássio (HUTTON et al., 1967; FARIAS & HERNÁNDEZ, 2017), promovem o aumento da fertilidade, melhorando assim o desenvolvimento de plantas (HALFFTER & EDMONS, 1982; GALBIATI et al., 1995). Além disso, são importantes no controle de moscas de importância médico- veterinária cujas larvas se desenvolvem em material em decomposição (FINCHER, 1981). Em relação ao processo de nidificação, as espécies são classificadas em três guildas funcionais: (I) tuneleiros (ou paracoprídeos), que são espécies que se alimentam e constroem seus ninhos em túneis logo abaixo do recurso alimentar, (II) os roladores (ou telecoprídeos) que rodam as esferas de alimento e nidificam longe do recurso alimentar e (III) os residentes que se alimentam e nidificam no próprio recurso (HALFFTER & EDMONDS, 1982). Deste modo, utilizam o mesmo recurso com estratégias diferentes, reduzindo a competição direta (HANSKI & CAMBEFORT, 1991) e resultando em variações comportamentais e ecológicas que se refletem em variações 27

morfométricas (HERNÁNDEZ et al., 2011). As diferenças morfométricas do corpo são relativamente evidentes entre os diferentes grupos funcionais de escarabeíneos: os endocoprídeos, como os integrantes do gênero Eurysternus são alongados no eixo anteroposterior e achatados no eixo dorsoventral; os paracoprídeos, como os do gênero Dichotomius, são ovalados com o élitro alargado lateralmente; os telecoprídeos, como os do gênero Canthon, são semelhantes aos paracoprídeos, porém são menos ovalados e levemente mais alongados no eixo anteroposterior (HERNÁNDEZ et al., 2011). Portanto, além de serem utilizados como indicadores ecológicos, os escarabeíneos são um grupo ideal para o estudo morfométrico, por possuírem diferentes guildas que resultam em uma clara distinção morfológica. Assim, a morfometria pode ser aplicada objetivando avaliar a relação da morfologia com a partilha de recursos, segregação de nicho, e por consequência a estruturação das comunidades (HERNÁNDEZ et al., 2011). Alterações na diversidade das comunidades resultam em alterações na diversidade morfométrica, pois espécies com características morfológicas únicas são perdidas. Tais perdas podem refletir em redução das funções ecossistêmicas prestadas pelos escarabeíneos, já que a eficiência na remoção e alocação de recursos está associada com o processo de nidificação (NERVO et al., 2014). Em escarabeíneos, as comunidades de matas adjacentes a cultivos transgênicos são estruturalmente diferentes daquelas comunidades adjacentes a cultivos convencionais (CAMPOS & HERNÁNDEZ, 2015a, CAMPOS & HERNÁNDEZ, 2015b) e escarabeíneos alimentandos com fezes contendo proteínas de milho transgênico apresentam alterações imunitárias e comportamentais (CAMPOS, 2016). Isto é uma evidência de que os escarabeíneos são responsivos à presença de OGMs, e que esses organismos transgênicos estão agindo como um filtro antrópico, afetando negativamente as comunidades e impactando nas funções ecossistêmicas dessas espécies.

USO DO SOLO A revolução neolítica (ocorrida há cerca de 10.000 anos) alterou o modo de vida dos humanos, deixando de serem nômades e começando a criar assentamentos e sociedades mais complexas. Desde então, o desenvolvimento da agricultura através da domesticação de plantas e animais influenciou a evolução humana, a transição do modo de vida de coletor-caçador para agricultor ofertou maiores quantidades de alimento, sustentando populações cada vez maiores. A importância da agricultura para a sociedade se estende desde o seu início até a sociedade moderna, 28

e a dependência humana por maiores quantidades de alimento é inegável. A necessidade de produção está associada ao aumento populacional, em 1965 a população mundial correspondia a cerca de 3 bilhões de habitantes, atualmente é estimada em 7 bilhões e a projeção para 2050 é de 9,7 bilhões, e questionamentos sobre como a produção agrícola irá suprir essa demanda estão sendo levantados (FAO, 2015). Projeções até 2050 calculam que a demanda da produção poderá aumentar entre 56% (em projeções mais otimistas) a até 96% (ELFERINK & SCHIERHORN, 2016). Por exemplo, em apenas 10 anos (2007-2017) a demanda mundial de produção de grãos aumentou mais de 20% de 2,100 a 2,600 milhões de toneladas (FAO, 2017). Além da maior demanda, as projeções dos efeitos das mudanças climáticas sobre a produção alimentar são preocupantes e estima-se que no Brasil, em importantes estados produtores de grãos (como o Mato Grosso), ocorra uma diminuição na produção de soja e milho em torno de 20% (COHN et al., 2016). Em uma visão conservacionista e sustentável, o aumento da produção deveria ser através da utilização de tecnologias mais eficientes (irrigação ou fertilizantes) nas áreas já cultivadas. Porém, frequentemente o aumento da produção está associado ao aumento das áreas de plantio ameaçando o pouco que resta das áreas preservadas (NOBRE et al., 2016). A ação antrópica praticamente já afetou todos os ecossistemas e o conceito de “natureza intocada”, referindo-se a áreas sem nenhuma influência humana, é uma utopia, sendo que contaminações de natureza antrópica já foram detectadas nos lugares mais remotos do planeta, como a Antártida onde foram encontradas resíduos do pesticida DDT em focas e pinguins (SLADEN et al., 1966), além de contaminação por plásticos e metais (SNAPE et al., 2001; TIN et al., 2009). O impacto humano na biodiversidade é praticamente uma marca registrada do Antropoceno devido à intensificação do uso da terra para diversas finalidades, promovendo um aumento exponencial nas taxas de extinção (entre 100 a 1000 vezes) e na degradação ambiental, ocasionando efeitos irreversíveis nos ecossistemas (CHAPIN et al., 2000). O estado de irreversibilidade do ecossistema à sua condição original, devido à ação antrópica, foi classificado como um “novo ecossistema” (MORSE et al., 2014). Este termo gerou muitas críticas e segundo alguns autores pode ser considerado apenas como um eufemismo para áreas degradadas (MURCIA et al., 2014). No Brasil, país considerado como o “celeiro do mundo”, cerca de 250 milhões de hectares são utilizados com finalidades agrícolas (IBGE, 2017). Segundo dados da FAO, o Brasil é o maior produtor de cana-de- 29

açúcar (736.101.487 t/ano), segundo maior produtor mundial de soja (86.760.520 t/ano), terceiro maior produtor de milho (79.881.614 t/ano), quinto maior produtor de algodão (2.669.161 t/ano) e o oitavo produtor mundial de arroz (12.175.602 t/ano). Associado com essa alta produção agrícola existe a supressão de florestas naturais: entre os anos 2000 e 2014 a área destinada ao uso agrícola no país aumentou 40% (cerca de 16 milhões de hectares) enquanto as áreas de mata nativa diminuíram 16% (cerca de 34 milhões de hectares) (IBGE, 2017). A conversão de áreas florestais em paisagens agrícolas está associada a uma simplificação dos ecossistemas, trazendo consequência direta à biota, afetando desde os indivíduos aos maiores níveis de organização ecológica como populações, comunidades e ecossistemas (MATSON et al., 1997; WILSON et al., 1999; HOLZSCHUH et al., 2006; HENDRICKX et al., 2007). A Mata Atlântica brasileira é o bioma mais afetado pela expansão agrícola, seguido do Cerrado e da Amazônia; essa pressão antrópica resultou na perda de aproximadamente 85% de sua cobertura original. Desse modo, a Mata Atlântica encontra-se altamente fragmentada, distribuída em pequenas manchas inseridas em uma matriz agrícola e nessa conformação de paisagem, as espécies ficam isoladas dentro de fragmentos. A capacidade dos organismos em se dispersar pela matriz agrícola é fundamental para evitar isolamento genético e consequente extinção local (KRAUSS et al., 2004; KIMATU, 2011). Comparando ao pool regional de espécies das áreas de mata, poucas espécies conseguem se estabelecer em áreas agrícolas, consequentemente as comunidades de áreas agrícolas costumam ser mais pobres do que as comunidades de vegetação nativa (SCHWEIGER et al., 2005). Dentre as unidades da Federação, Santa Catarina possui a maior taxa de preservação da Mata Atlântica (cerca de 20%). Porém em 2014 foram desmatados 67.348 km2 de cobertura original (IBGE, 2015). Em detrimento do desmatamento ocorreu um aumento nas áreas de produção agrícola: em 2014 cerca de 30 milhões de hectares foram utilizados para produção de soja, 16 milhões para produção de milho, 3 milhões para produção de feijão e 2 milhões para produção de arroz (EPAGRI, 2015). Santa Catarina é o oitavo maior produtor nacional de milho, correspondendo com 2,9% da produção nacional (IBGE, 2015), atualmente cerca de 400 mil hectares são utilizados para cultivo de diferentes tipos de milho e estima-se que a produção em 2017 alcance 3,06 milhões de toneladas (EPAGRI, 2017). A região de São Miguel do Oeste (região amostrada nesse estudo), apresentou o maior crescimento na produção de milho em Santa Catarina e é estimado um aumento em 30

30% na safra de 2017 (EPAGRI, 2017). Além de ser uma importante região produtora, o Oeste Catarinense também se destaca no cultivo de variedades locais, sendo que o município de Anchieta (também abrangido nesse estudo) é considerado como capital estadual do milho crioulo, com cerca de 20 variedades produzidas (VOGT et al., 2007). As plantações com práticas tradicionais, como o caso do milho crioulo, permitem que variedades sejam selecionadas em um processo de melhoramento realizado ao longo de gerações pelos próprios agricultores, os quais detêm elevado conhecimento desta prática (ABREU et al., 2007). Os cultivos crioulos são importantes para os pequenos agricultores, por serem de baixo custo de implantação e são utilizados na alimentação, na cultura e também como fonte de renda (CATÃO et al., 2013). O milho crioulo possui elevado potencial de adaptação em condições ambientais específicas (PATERNIANI et al., 2000) e as populações crioulas são muito importantes por constituírem fonte de variabilidade genética, que podem ser exploradas na busca por genes tolerantes a condições adversas (ARAÚJO & NASS, 2002). Além disso, estas variedades são mais resistentes a doenças e pragas (CARPENTIERE-PÍPOLO et al., 2010) e no seu manejo a utilização de inseticidas, herbicidas e fungicidas é muito baixa, ou inexistente, sendo menos agressivo ao ambiente. As motivações para a manutenção das variedades crioulas são por questões culturais (tradição, sabor e beleza) e econômicas, como a redução nos custos de produção, adaptação às condições climáticas e o maior rendimento (VOGT et al., 2007). A produção de variedades locais é voltada principalmente às necessidades das propriedades, como a alimentação animal e produção de alimentos para consumo humano, o excedente é comercializado para a fabricação de ração animal (VOGT et al., 2007). Apesar de existirem incentivos por parte dos órgãos estaduais para o cultivo de variedades crioulas, a produção do milho crioulo está perdendo espaço para variedades convencionais e transgênicas. Por exemplo, na região de estudo deste trabalho, em áreas onde o milho crioulo era cultivado desde a década de 1960 foram substituídos por variedades transgênicas na safra de 2016/2017 (observação pessoal). O cultivo realizado com variedades de milho convencional utiliza em grandes quantidades inseticidas com base em piretróides e organofosforados, pois a implementação em larga escala dos cultivos, com a consequente supressão das áreas de floresta nativa, propiciou a infestação por insetos fitófagos de importância econômica. O uso de inseticidas na agricultura aumenta a produtividade, mas também aumenta a concentração destes compostos nos alimentos e no ambiente, 31

causando efeitos negativos na saúde humana (ANDERSSON et al., 2014). Também agem intensamente contra insetos não-alvo que possuem funções-chave dentro dos ecossistemas (e.g. polinizadores e detritívoros) (INGHAM, 1985; INGLESFIELD, 1989). Objetivando a redução do uso de inseticidades, com o avanço das técnicas da engenharia genética, em meados de 1980, começaram a serem desenvolvidas as plantas geneticamente modificadas. O desenvolvimento de plantas transgênicas tem como base pesquisas antigas: em 1911, Ernst Berliner isolou uma bactéria responsável pela morte de lagartas de Anagasta kuehniella e como homenagem a província de Turingia na Alemanha ele denominou a bactéria como Bacillus thuringensis (Bt). Na década de 20 o Bt começou a ser utilizado como pesticida na Europa e em meados da década de 30 a França começou a comercializar formulações baseadas em Bt. As plantas transgênicas expressam as toxinas de Bt (principalmente Cry-toxinas) com o objetivo de eliminar populações de insetos de importância econômica, principalmente lepidópteros e coleópteros. A vantagem da utilização de transgênicos seria a diminuição na demanda na utilização de inseticidas e a sua aparente segurança, já que algumas toxinas são hospedeiro-específica devido a particularidades da fisiologia do inseto- alvo (HOFMANN et al., 1988). Porém, mesmo nos cultivos transgênicos são utilizados inseticidas, e já foram relatados efeitos negativos dos transgênicos nas comunidades de artrópodes (MARVIER et al., 2007), acúmulo de toxinas transgênicas na cadeia trófica (ALVAREZ et al., 2008), contaminação em corpos d‟água (AXELSSON et al., 2011), resistência dos insetos alvos (GASSMANN et al., 2011) e efeitos letais e subletais em organismos não alvos como abelhas (HAN et al., 2010) e escarabeíneos (CAMPOS, 2016). O conhecimento dos efeitos dos organismos geneticamente modificados (OGM) na biodiversidade está muito aquém da velocidade em que esses OGMs estão sendo implementados. Em vinte anos (1996-2016) a área mundial de produção de transgênicos saltou de 1,6 para 185 milhões de hectares e o Brasil é o segundo maior produtor mundial de transgênicos com 49 milhões de hectares (distribuídos em soja, milho e algodão), respondendo por 27% da produção mundial (ISAAA, 2016) e por 17% da área mundial de cultivos transgênicos (JAMES, 2010). Nos cultivos brasileiros cerca de 90% da soja, 88% do milho e 78% do algodão são geneticamente modificados (ISAAA, 2016). 32

HIPÓTESES

A tese está dividida em quatro capítulos derivados de amostragens de escarabeíneos realizadas na região de São Miguel do Oeste (26°43'31''S, 53°31'05''O), no extremo oeste de Santa Catarina, em uma paisagem agrícola com remanescentes de Mata Atlântica. O primeiro capítulo baseia-se na hipótese de que cultivos de milho com organismos geneticamente modificados, para o controle de insetos que atacam os cultivos (principalmente lepidópteros), afetam negativamente as comunidades de escarabeíneos dos próprios cultivos e/ou das matas nativas adjacentes, resultando em um empobrecimento das comunidades devido à perda de espécies. No segundo capítulo trabalhamos com a hipótese de que as espécies presentes no milho transgênico possuem alterações em seu fitness devido ao acúmulo das proteínas transgênicas na cadeia trófica, que afetariam as populações de escarabeíneos através da alimentação de fezes ou carcaças contaminadas com proteínas transgênicas. Para uma espécie de escarabeíneo (Canthon quinquemaculatus) utilizamos a forma do corpo como um indicativo do seu fitness e, através de técnicas de morfometria geométrica, comparamos a forma do corpo entre indivíduos presentes nos cultivos e indivíduos das matas adjacentes, objetivando observar possíveis efeitos dos transgênicos em uma espécie não-alvo. No terceiro capítulo nos baseamos na hipótese de que o nicho das espécies pode ser descrito de um ponto de vista morfométrico através das métricas do hipervolume, e que as comunidades das áreas de cultivo de milho possuem menor diversidade morfométrica em comparação às comunidades das matas nativas. Também hipotetizamos que as comunidades dos cultivos possuem maiores buracos no hipervolume morfométrico, demonstrando a extinção local de escarabeíneos nas áreas agrícolas. No quarto capítulo, durante o período de doutorado sanduíche no Museu de Ciências Naturais de Madri, realizamos testes espectrofotométricos nos élitros de cinco espécies de escarabeíneos distribuídos ao longo da paisagem agrícola. A hipótese é de que as espécies que habitam, preferencialmente, os cultivos agrícolas possuem diferentes características fotobiológicas em relação às espécies que habitam as áreas de mata, já que nos cultivos não há atenuação da 33

radiação solar, o que pode ser um fator limitante para as espécies na colonização dessas áreas.

OBJETIVOS

OBJETIVO GERAL Esse trabalho teve como objetivo geral avaliar os efeitos do uso do solo (florestas nativas e cultivos agrícolas) sobre os besouros escarabeíneos. Inicialmente objetivou-se avaliar se existe um efeito dos cultivos transgênicos na diversidade (nível de comunidade) e na morfometria – representando o fitness - (nível populacional) de uma espécie de escarabeíneo. Posteriormente objetivou-se avaliar se há efeito dos cultivos na diversidade morfométrica das comunidades e nas características espectrofotométricas das espécies.

OBJETIVOS ESPECÍFICOS

Artigo I: Dung beetles and the conservation of diversity in an agricultural landscape with maize fields and Atlantic Forest remnants: the influence of creole, conventional and transgenic crops.

Objetivo: Avaliar os possíveis impactos dos cultivos de milho convencional, crioulo e transgênico tanto nas comunidades de escarabeíneos dos próprios cultivos quanto das comunidades dos remanescentes de Mata Atlântica adjacentes aos cultivos. Além do tipo de milho também consideramos como variáveis explicativas a quantidade de insumos utilizados nos cultivos, as características da vegetação e a distância espacial entre as áreas.

Artigo II: Morphometric modifications in Canthon quinquemaculatus Castelnau, 1840 (Coleoptera: Scarabaeinae): sublethal effects of transgenic maize?

Objetivo: Avaliar se a espécie de escarabeíneo Canthon quinquemaculatus apresenta diferenças morfométricas quando presente em cultivos de milho convencional, crioulo ou transgênico em relação aos indivíduos das populações das florestas nativas adjacentes.

Artigo III: Local extinctions may be evidenced by the holes of the morphometric hypervolume in dung beetles communities

Objetivo: Descrever morfometricamente as comunidades de escarabeíneos e, através das métricas do hipervolume, avaliar a relação da diversidade morfométrica e do tamanho dos buracos do hipervolume com a diversidade taxonômica. Assim, objetivou-se comparar a diversidade morfométrica e o tamanho dos buracos entre as comunidades da Mata Atlântica e dos cultivos agrícolas.

Artigo IV: Elytra absorb ultraviolet radiation but transmit infrared radiation in Neotropical Canthon species (Coleoptera, Scarabaeinae)

Objetivo: Determinar a capacidade dos élitros de cinco espécies de escarabeíneos em refletir, transmitir e absorver radiações em um amplo espectro eletromagnético: ultravioleta, visível e infravermelho, para avaliar se há relação entre as características espectrofotométricas com a ocupação do hábitat, pelas espécies em áreas abertas (cultivos de milho) ou em áreas florestais.

ARTIGO I: DUNG BEETLES AND THE CONSERVATION OF DIVERSITY IN AN AGRICULTURAL LANDSCAPE WITH MAIZE FIELDS AND ATLANTIC FOREST REMNANTS: THE INFLUENCE OF CREOLE, CONVENTIONAL AND TRANSGENIC CROPS

Victor Michelon Alves1, Eduardo Luís Hettwer Giehl1, Paulo Emilio Lovato2, Fernando Zagury Vaz-de-Mello3, Marcelo Betancur Agudelo2, Malva Isabel Medina Hernández1

1 Programa de Pós-Graduação em Ecologia, Departamento de Ecologia e Zoologia, Universidade Federal de Santa Catarina, Florianópolis, Santa Catarina, 88040–900

2 Programa de Pós-Graduação em Recursos Genéticos Vegetais, Centro de Ciências Agrárias, Universidade Federal de Santa Catarina, Florianópolis, Santa Catarina, 88040–900

3 Programa de Pós-Graduação em Zoologia, Departamento de Biologia e Zoologia, Instituto de Biociências, Universidade Federal de Mato Grosso, Cuiabá, Mato Grosso, 78060-900

Artigo submetido ao periódico Agriculture, Ecosystems & Environment

Dung beetles and the conservation of diversity in an agricultural landscape with maize fields and Atlantic Forest remnants: the influence of creole, conventional and transgenic crops

Abstract Land use change is a global threat to biodiversity. Besides converting natural landscapes into monocultures, the use of genetically modified organisms is increasing, even though their effects on natural assemblages are still poorly known. The possible effects of transgenic crops on ecosystems can be studied using ecological indicators, such as dung beetles. We assessed the effect of creole, conventional, and transgenic maize crops on dung beetle assemblages in a landscape that also contained forest fragments in southern Brazil. We established 15 blocks, five on each type of maize fields. Each block consisted of a landscape with native forests and maize fields. A total of 2126 dung beetles were captured, belonging to 38 species. The forest-crop landscape associated with creole maize had a greater richness of dung beetle species. Conversely, an impoverished assemblage was detected in the transgenic maize crop landscape. The reduction of indicator insect diversity in landscapes influenced by transgenic crops suggests that the type of maize fields influences biodiversity conservation. In addition, the negative effects of transgenic crops went beyond maize fields, reaching native forest fragments adjacent to crops.

Keywords: Agrobiodiversity; Bioindicators; Coleoptera; Ecology; Scarabaeinae; Zea mays.

Introduction Deforestation and conversion of tropical forests to pasture and agriculture has changed landscapes, causing negative effects on biodiversity (Chapin et al., 2000). Such land use changes are generally associated with habitat loss (Tilman et al., 2001), and most of the species is intolerant to the modifications caused by anthropic action (Turner 1996), the inability to maintain viable populations in agricultural areas results in species extinction (Turner 1996; Brooks et al., 2000,). Maize or corn (Zea mays L.) cultivation began around 6000 years ago in southern Mexico (Sluyter and Dominguez, 2006) and now is cultivated worldwide and represents one of the most important global crops. Together with rice and wheat, maize accounts for 94% of the world‟s consumption of cereals by humans (Ranum et al., 2014). Brazil is the third largest producer of maize, harvesting 8% of the global production (Ranum et al., 2014). In Brazil, usually, three types of maize are cultivated: (1) creole or landrace, (2) conventional and (3) transgenic. Creole seeds are selected over generations by the farmers themselves (Abreu et al., 2007). Creole maize is more resistant to pathogens and pests (Carpentieri-Pípolo et al., 2010), and acts as a source of genetic variability that can be screened for tolerance to adverse conditions (Araújo and Nass, 2002). For farmers employing low technology, creole varieties can perform equally well or even better than conventional varieties (Carpentieri-Pípolo et al., 2010). On extensive production systems, conventional maize is normally more productive because of the use of insecticides with pyrethroids and organophosphates. However, the concentration of these compounds both in food and the environment are causing adverse effects on human health (Tago et al., 2014) and showing side effects on insects with key ecosystem functions (e.g. pollinators and detritivores) (Palmquist et al., 2012). Advances in genetic engineering to increase production and suppress the use of insecticides lead to the develop of genetically modified (GM or just transgenic) crops. Transgenic maize expresses toxins from Bacillus thuringiensis (primarily Cry-toxins) aiming at damage control caused by insects of economic importance, especially Lepidoptera and Coleoptera. Cultivation of GM crops has increased worldwide, in 2016 the global area cultivated with GM maize reached 60 million hectares and continues to grow in many countries, including Brazil (ISAAA, 2016). The advantage of using GM crops to decrease the use of insecticides has been questioned, and several studies have shown the development of resistance by target insects (Gassmann et al., 2011, 39

2014). In addition, negative effects of GM crops on non-target organisms, like butterflies and dung beetles, have already been found (Losey et al., 1999; Campos and Hernández, 2015a). Although the mechanistic effects of the transgenics on non-target species are unclear, morphometric alterations have been observed in dung beeltes inhabiting transgenic maize crops (Alves and Hernández, 2017). These morphological changes may be indicative of a possible impact of transgenic crops on the species physiology, resulting in long-term mortality (Campos and Hernández, 2015a). Furthermore, alterations in abundance and structure of dung beetles communities have been described in native forests adjacent to transgenic crops (Campos and Hernández 2015a, 2015b). The study of ecological indicators is a straightforward way to assess the effects of crops on ecosystems, showing measurable characteristics of structure, composition and ecological functions. Among such organisms, insects stand out because of their high species richness, abundance, biomass, and key ecosystem functions (Niemi and McDonald, 2004). Dung beetles are widely used as ecological indicators because of both ease and cheap sampling (Gardner et al., 2008a) and from being responsive to land use change (Halffter and Favila, 1993; review in Nichols et al., 2007; Gardner et al., 2008b). Dung beetles play important ecosystems functions, especially in nutrient cycling by being coprophagous and showing the behavior of burying organic material in the soil (Halffter and Matthews, 1966). Therefore, dung beetles increase soil fertility by replacing soil nutrients and thus improving plant growth (Halffter and Edmonds, 1982). Across agricultural landscapes, dung beetle assemblages can be seen as a metacommunity (Leibold, 2004). Forest fragments are interspersed within the agricultural matrix, creating mosaics in the landscape, with permeability of the agricultural matrix allowing interpopulation gene flow (Vandermeer, 2011). Moreover, the type of crop in a landscape can influence the survival and dispersal ability of species. Consequently, changes in dung beetle richness and abundance may be either a direct product of the type of crop they live in, or an indirect product, when it affects forest assemblages adjacent to crops. Dung beetle assemblages living in areas of maize cultivation may consist of species adapted to open areas and species coming from forest areas adjacent to crops, which show up in crops as transitory. Therefore, vegetation structure and conservation status of forest fragments may be an important driver of dung beetle assemblages, although the type of maize cultivated may also affect the dung beetle community. 40

We expect that landscapes in which creole corn is cultivated show a greater diversity and abundance of dung beetles than the areas under the influence of transgenic and conventional crops. Specifically, we expect that landscapes with transgenic and conventional crops result in the loss of dung beetle species because of the sensitivity to the treatment itself or to the environmental changes that it causes. We also expect that the effects mentioned above are not only limited to the extent of fields areas, resulting in distinct side effects on the dung beetle assemblages found in forests adjacent to different maize crop types. In this context, we aimed to assess the effect of the type of cultivated maize (conventional, creole, or transgenic) on the structure of dung beetle assemblages, considering both the assemblages established in the crops, as well as those within forest fragments adjacent to crops.

Materials and Methods Study area and description of the vegetation structure of forest fragments The study was carried out in the region of São Miguel do Oeste, state of Santa Catarina, Brazil (26°43'31''S, 53°31'05''W). Altitude ranges from 440 up to 760 m. The region has a humid subtropical climate (Cfa) under the Köppen climate classification. Annual mean temperature is between 16.3º C and 17.9º C and annual precipitation between 1790 mm and 2280 mm (Thomé et al., 1999). The landscape is a mosaic of corn crops and Atlantic Forest fragments (Seasonal Forests) with high tree diversity and dense canopy cover. These fragments cover a small percentage of original forest that must be maintained on farms because their protection is now enforced by the law (Silva and Raniere, 2014). However, along the recent history of land use change, these fragments helped to maintain around 16% of the original landscape (Vibrans et al., 2013). Samples were collected in 15 blocks that were placed at least 1 km apart. Each block was a landscape with two components – native forest and maize fields, totaling 30 samples (Fig. 1; see Fig.S1, S2 and S3 for landscape details). Blocks were replicated five times on every one of conventional, creole, and transgenic maize fields (Fig. S4). To assess the diferent kind of management, questions were asked about farming practices with all farmers in the 15 properties. The questions were about the maize variety, and the use of insecticide, herbicide, and fungicide. With the answers, the intensity of the use of inputs in each area was calculated (Table S1).

Figure 1. Location of dung beetle samples in the region of São Miguel do Oeste, state of Santa Catarina, Brazil. Each site (points magnified in the figure) represents a block of samples consisting of two landscape units: maize fields and adjacent forest fragments. Blocks were replicated five times for each type of maize grown in the landscape: conventional, creole and transgenic

To assess the vegetation structure on native forests, two 100 m2 plots were established in every block and placed 10 m apart from each other. In each plot, all the trees with a diameter at breast height > 5 cm were identified, and their diameter and height were measured. All tree species were identified and classified into either pioneer, light dependent, or shade tolerant following specific literature (e.g. Ferreira et al., 2013, Scipioni et al., 2013). Ground cover were measured in four 1 m² sub-plots, which were placed either to the northeast, southeast, southwest, and northwest and one meter away from plot corners. For ground cover, percentages of litter coverage, green cover (soil covered with either grass, bushes, or tree seedlings and saplings) and bare ground (no vegetation or litter) were visually estimated in sub-plots. Canopy cover was estimated in each the four above mentioned directions, with the aid of 10 × 10 cm squared paperboard at 40 cm from the observer, at a 20° angle to the zenith.

Dung beetle sampling in maize fields and adjacent forest fragments Dung beetles were collected during the summer, from December 2014 to February 2015. At that time, maize fields were between 60 and 90 days of development. Dung samples were taken in both forest fragments and maize fields using pitfall traps, because such traps are the most efficient in sampling soil insects (Lobo et al., 1988). Pitfall traps were made with 15 cm × 20 cm (diam. x depth) plastic containers and buried in the ground. Each pitfall was protected by a plastic lid with a bait hung in a package made of voile fabric. A water-detergent solution was added to pitfall traps to catch insects. Five sampling points were established within each landscape unit (forest and maize fields) and spaced by 10 m. Each sampling point consisted of a pair of traps placed five meters away from each other. Out of each pair of traps, one was baited with 10 g of human feces, and the other baited with 10 g of carrion pork, totaling ten traps per landscape unit (Fig. S5). Thus, the sampling design spanned over 15 landscapes, each with two landscape units (one of forest and one of maize) and ten traps, totaling 300 traps. All traps were open for 48 hours and the captured insects were fixed in 70% alcohol.

Data Analysis To compare the richness found across landscape types, the confidence interval (95%) of the observed richness was calculated using the software EstimateS v.9 (Colwell, 2013). To assess sufficiency of the sampling effort, species richness was estimated using the Chao1 index in EstimateS v.9 (Colwell, 2013) and the sample coverage was calculated using the entropart package in R software (R Core Team, 2015). A dissimilarity matrix was constructed using the Bray-Curtis index. Before that, species abundance data were fourth-root transformed to reduce the influence of abundant and common species. The dissimilarity matrix was analyzed by non-metric multidimensional scaling analysis (n-MDS). In addition, similarity percentages analyses (SIMPER) between landscape units types were calculated with Bray- Curtis dissimilarities. This set of analyses was calculated using the vegan package in R (R Core Team, 2015). The environmental structure and vegetation complexity of forest fragments was first assessed visually with a principal component analysis (PCA). Several variables describing tree structure and soil cover were included in the PCA to search for patterns across distinct landscapes. Next, a multivariate analysis of variance (MANOVA) was 43

applied to test for significant differences among landscapes. To describe the effect of spatial location of sites, a matrix of spatial vectors was calculated using the method of principal coordinates of the truncated geographic distance (PCNM) between sites. The PCA, MANOVA, and PCNM were calculated in R (R Core Team, 2015), to PCA and PCNM was utilized the vegan package. To test which explanatory variables are affecting the structure of dung beetle assemblages, a nested generalized linear model for multivariate data (GLMmv) was carried out, with negative binomial distribution with a log link. In this analysis, the dung beetle abundance data were used as dependent variables and the maize abundance data was nested in the forest areas. Six explanatory variables were used: 1) landscape unit (maize crops or native forest); 2) maize type; 3) agricultural inputs use; 4) location (PCNM1); 5 and 6) vegetation structure (PC1 and PC2: the first and second principal components of the PCA on vegetation structure data). The GLMmv was calculated with the mvabund package (Wang et al. 2012) in R (R Core Team, 2015). Model residuals were checked for homoscedasticity (Fig. S6).

Results

Dung beetle diversity A total of 2126 dung beetles were collected, belonging to fifteen genera and 38 species (Table 1). Out those species, 14 were found exclusively in forest fragments: Canthidium cavifrons, Canthidium sp.2, Canthon laminatus, Canthon lividus lividus, Canthon aff. luctuosus, Canthonella catharinensis, Canthonella aff. instriata, Deltochilum brasiliense, Deltochilum morbillosum, Deltochilum rubripenne, Dichotomius sp., Eurysternus caribaeus, Scybalocanthon nigriceps and Uroxys dilaticollis, and 11 were exclusive for maize fields: Anomiopus sp., Canthidium moestum, Canthidium sp.3, Canthon podagricus, Canthon tetraodon, Canthon aff. lituratus, Dichotomius bicuspis, Eurysternus aeneus, Pseudocanthon sp., Sulcophanaeus menelas and Trichillum externepunctatum.

Table 1. Dung beetle species sampled in 30 landscape units (15 maize fields and 15 nearby forest fragments) and where three types of maize are grown (CO: Conventional; CR: Creole; TR: Transgenic) in the region of São Miguel do Oeste, state of Santa Catarina, Brazil.

Forest fragments Maize fields Species CO CR TR Total CO CR TR Total Anomiopus sp. 0 0 0 0 0 0 1 1

Canthidium cavifrons 0 3 1 4 0 0 0 0 Balthasar,1939 Canthidium dispar Harold, 4 8 1 13 3 0 8 11 1867

Canthidium moestum Harold, 0 0 0 0 3 11 3 17 1867 Canthidium aff. trinodosum 29 33 0 62 20 55 46 121 (Boheman, 1858) Canthidium sp. 1 3 0 0 3 16 0 0 16 Canthidium sp. 2 2 0 1 3 0 0 0 0 Canthidium sp. 3 0 0 0 0 0 0 1 1 Canthon angularis Harold, 35 56 0 91 1 0 0 1 1868 Canthon chalybaeus 13 8 0 21 114 67 27 208 Blanchard, 1846 Canthon laminatus 0 1 0 1 0 0 0 0 Balthasar, 1939 Canthon lividus lividus 14 4 0 18 0 0 0 0 Blanchard, 1845 45

Canthon podagricus 0 0 0 0 2 3 0 5 Harold, 1868 Canthon quinquemaculatus 72 35 67 174 12 17 28 57 Castelnau, 1840

Canthon tetraodon 0 0 0 0 0 9 0 9 Blanchard, 1845 Canthon aff. lituratus 0 0 0 0 0 1 0 1 (Germar, 1824) Canthon aff. luctuosus Harold, 24 16 59 99 0 0 0 0 1868 Canthon aff. mutabilis Lucas, 1 0 0 1 0 0 7 7 1859 Canthonella catharinensis 9 5 1 15 0 0 0 0 (Pereira & Martínez, 1956) Canthonella aff. instriata 0 0 3 3 0 0 0 0 (Boucomont, 1928) Coprophanaeus saphirinus 48 85 38 171 1 3 6 10 (Sturm,1826) Deltochilum brasiliense 76 69 41 186 0 0 0 0 (Castelnau,1840) Deltochilum morbillosum 37 49 49 135 0 0 0 0 Burmeister, 1848 46

Deltochilum rubripenne 4 1 0 5 0 0 0 0 (Gory, 1831) Dichotomius bicuspis 0 0 0 0 1 0 0 1 (German, 1824) Dichotomius nisus (Olivier, 2 0 0 2 13 30 0 43 1789)

Dichotomius sericeus (Harold, 63 184 143 390 2 2 10 14 1867) Dichotomius sp. 0 1 0 1 0 0 0 0 Eurysternus aeneus Génier, 0 0 0 0 8 0 0 8 2009 Eurysternus caribaeus 10 1 9 20 0 0 0 0 (Herbst,1789)

Eurysternus parallelus 25 4 49 78 3 0 4 7 Castelnau,1840

Ontherus sulcator 0 1 0 1 0 2 0 2 (Fabricius, 1775)

Onthophagus aff. hirculus 2 0 0 2 14 11 14 39 Mannerheim, 1829

Pseudocanthon 0 0 0 0 19 9 3 31 sp.

Scybalocanthon nigriceps 0 3 0 3 0 0 0 0 (Harold, 1868) 47

Sulcophanaeus menelas 0 0 0 0 1 1 0 2 (Castelnau, 1840) Trichillum externepunctatum 0 0 0 0 1 8 0 9 Preudhomme de Borre, 1886 Uroxys dilaticollis 0 3 0 3 0 0 0 0 (Blanchard, 1845) Total abundance 473 570 462 1505 234 229 158 621

Regarding total abundance, 1505 dung beetles were collected in forest fragments, more than twice that found in maize fields, with 621 individuals (Table 2).

Table 2. Abundance, observed Richness (with confidence intervals of 95%), estimated Richness Chao-1 and Sample coverage of dung beetle assemblages in forest fragments and cultivation areas with three different maize types: Conventional (CO), Creole (CR) and Transgenic (TR), collected in the region of São Miguel do Oeste, State of Santa Catarina, Brazil. Measurements of dung beetle Forest Fragments Maize fields assemblages CO CR TR CO CR TR Abundance 473 570 462 234 229 158

20 21 13 18 15 13 Richness (17.26 - (17.41 (8.69- (14.25 – (13.46 – (10.4- (C.I. 95%) 22.74) - 24.6) 17.31) 21.55 16.54) 15.5)

Richness estimator 20.0 31.0 19.0 21.3 15.3 14.0 Chao-1 (sample (87%) (89%) (92%) (76%) (85%) (74%) coverage)

Analyzing the dung beetles richness of forests, a poorer assemblage was found in forests adjacent to transgenic maize fields (Table 2). Richness difference was significant between the forests adjacent to the creole (21 species) and transgenic (only 13 species) maize crops, with a small overlap between conventional (20 species) 48

and transgenic maize crops (Table 2). Although similar results was observed within maize fields (13 species in transgenic maize fields, 18 and 15 species in conventional and creole maize fields respectivelly), the difference was not significant because of overlapping confidence intervals (Table 2). The richness estimator Chao 1 evidences the richness difference between the assemblages from forests adjacent to creole (31 species) and transgenic crops (19 species). This diversity loss is reliable based on high sample coverage, around 90% for both. For all trataments sample coverage was high, among 74 and 92% of the richness estimated by Chao-1, indicating sample adequacy, mainly in forest areas (Table 2). The key changes in species composition occur from forest fragments to maize fields, therefore being represented in the first axis of the NMDS ordination of the species abundance matrix (Fig. 2). Assemblages within maize fields differed along the second ordination axis and indicated larger differences from conventional to transgenic and creole maize fields. Composition of assemblages within forest fragments overlapped more, yet with some lower similarity of fragments in conventional maize landscapes towards the other landscape types (Fig. 2).

Fig. 2 Non-metric multidimensional scaling (NMDS) ordination plot for dung beetle assemblages sampled in 30 landscape units (15 maize fields and 15 nearby forest fragments) and where three types of maize are grown in the region São Miguel do Oeste, state of Santa Catarina, Brazil.

Changes in dung beetle species distribution and abundance across landscapes were large: the average dissimilarity between assemblages found in forests was 61% and 73% between assemblages found within crops (SIMPER tests). The most representative species for each area (up to 60% contribution to dissimilarities) are shown in Fig. 3 for forest fragments and Fig. 4 for maize fields.

Fig. 3 Dissimilarity of dung beetle assemblages occurring in forest fragments nearby maize fields where three types of maize are grown (conventional, creole and transgenic) in the region of São Miguel do Oeste, state of Santa Catarina, Brazil. Species drawn are the most representative in each landscape unit and account for ~ 60% of the differences between them.

Fig. 4 Dissimilarity of dung beetle assemblages occurring in fields where three types of maize are grown (conventional, creole and transgenic) in the region of São Miguel do Oeste, state of Santa Catarina, Brazil. Species drawn are the most representative in each landscape unit and account for ~ 60% of the differences between them.

Variables affecting the structure of dung beetle assemblages The descriptions of the vegetation structure of forest fragments nearby maize fields included a total of 492 trees belonging to 81 species (Table S2). The environmental variables of fragments, including vegetation structure and canopy and soil cover, was unrelated to the type of cultivated maize (MANOVA, Pillai‟s Trace = 1.45, F = 1.49, P = 0.29). In the PCA, the first axis (PC1) explained 46% of the observed variation and was associated with sites with greater litter cover and small amounts of exposed soil (Table 3). The second axis (PC2) explained 25% of changes in vegetation structure and was associated to light-dependent tree species and soil green cover (Table 3).

Table 3. Scores of the first two axes of the principal component analysis (PC1 and PC2) of the variables describing vegetation structure of the forest fragments nearby maize fields in the region of São Miguel do Oeste, state of Santa Catarina, Brazil. Tree functional groups area as follow: pioneer (P), secondary light-demanding (SL), and secondary shade tolerant tree species (SS).

Variable PC1 PC2 Soil green cover 1.244 1.718 Soil litter cover 4.150 -1.546 Bare ground cover -4.187 0.032 Canopy cover 1.700 -1.368 Canopy height 0.168 0.178 Tree basal área 0.001 0.004 Number of SL species -1.668 -3.991 Number of P species 0.244 0.313 Number of SS species 0.182 -0.105

The results of the GLMmv, including both PCA axis and geographic location, showed a strong effect of land use, since the dung beetles assemblages in maize crops are different from those in the forest (Table 4). Moreover, the analysis evidenced that the type of corn grown in the landscape (conventional, creole, or transgenic) and the use of agriculture inputs also influences the structure of dung beetle assemblages. This effect occurs within maize fields and reaches adjacent forest areas (Table 4).

Table 4. Explanatory variables used in the generalized linear model for multivariate responses (GLMmv) to explain differences in composition and abundance patterns of dung beetle assemblages in the region of São Miguel do Oeste, state of Santa Catarina, Brazil. Wald Variables P score Intercept 7.52 0.001 Landscape: Creole (Compared to Conventional) 5.77 0.005 Landscape: Transgenic (Compared to 5.01 0.020 Conventional) Maize: Conventional (Compared to the 10.26 0.001 corresponding Forest) Maize: Creole (Compared to the corresponding 9.57 0.001 Forest) 53

Maize: Transgenic (Compared to the 8.55 0.001 corresponding Forest) Agricultural inputs 8.36 0.003 PC1 (representing mostly soil coverage) 7.41 0.001 PC2 (representing mostly vegetation structure) 10.58 0.001 PCNM1 (geographical location) 6.31 0.001

Discussion Our results show that richness, abundance, and composition of dung beetle assemblages changed as result of differences across landscapes. Such changes are either drastic in the case of native forest fragment conversion to monocultures, or subtler, when contrasting distinct types of maize fields. There was an effect of anthropic filters, because the compositional differences found in the NMDS were consistent even after considering environmental variables and geographic location of sites. This results points the importance of forest fragments in the maintenance of regional species pools, which ensures local biodiversity with forest fragments mostly acting as sources for cultivation fields. The decrease in richness and abundance due to the conversion of tropical forests into agricultural landscapes or monocultures is well documented for dung beetles, and may even decrease the species richness by half when comparing either pastures (Horgan, 2008) or eucalyptus plantations to nearby forests (Gardner et al., 2008b). Smaller decreases in species richness have been reported from converted landscapes to open vegetation (Estrada et al., 1998). Even though land conversion to agriculture negatively impacts dung beetle assemblages (review in Nichols et al., 2007), some types of crops (e.g. creole crop) can mitigate biodiversity loss. Here, landscapes with creole maize fields showed at least slightly higher dung beetle species richness. Therefore, landscapes with creole or conventional maize fields can maintain higher biodiversity than landscapes with transgenic types maize fields. Simper analysis showed that the largest species are in creole maize (e.g. Dichotomius nisus). Large nocturnal tunneler are important in ecosystem services, because may account for up to 75% of the organic matter removal (Slade et al. 2007). The tunneler D. nisus is one of the main species that contribute in organic matter removal in agricultural systems in Southern Brazil (Farias and Hernández, 2016). The same analysis showed that larger species also are related to forest fragment adjacent to creole maize 54

crops. Therefore, the largest species presence in creole maize landscape (crop and adjacent forest), shows the importance of this type of maize in the maintenance of the biodiversity and ecosystem services performed by these insects. Dung beetle assemblages in the transgenic crop landscape had the lowest richness and abundance, which corroborates previous concerns that transgenic crops can affect the whole arthropod community (Marvier et al., 2007). The effect of transgenic maize on insect species that inhabit crops may increase mortality through the food chain, such as the case of the predatory lacewing, Chrysoperla carnea: individuals of C. carneae showed an increased mortality when fed Lepidoptera reared on Bt-maize (Dutton et al., 2002). Moreover, transgenic crops can be a risk to the health of immature stages of non-target Lepidoptera (Losey et al., 1999). For dung beetles, Campos and Hernández (2015b) observed changes in assemblages found in forests nearby transgenic maize fields. Moreover, they found that these areas had greater abundance of two species of the Eurysternus genus than areas where maize fields were from the conventional type. Since Eurysternus spp. are considered inferior competitors because of their feeding and reproductive behavior (Scholtz et al., 2009), its increased abundance nearby transgenic maize fields points to a decrease in stronger competitors within the assemblage. Here, we corroborate this pattern, finding Eurysternus parallelus with higher abundance in forests within landscapes with transgenic maize fields. In addition, a reduction in dung beetle abundance from forest fragments nearby conventional to transgenic maize fields has been reported by Campos and Hernández (2015a). The authors attributed the reduction to the combination of transgenic seeds and maize management, accelerating the process of loss of diversity. Here, we extended the framework and assessed dung beetle assemblages in maize fields and nearby forests and corroborate the pattern, with loss of diversity in landscapes with transgenic fields. In relation to management we observed a high use of agricultural inputs in conventional and even in transgenic crops (Table S1). Therefore, the mechanistic links that underpin the loss of diversity may be an association between transgenic proteins and agricultural inputs. In this context creole maize stands out due to its very low input utilization, consequently with less impact on the environment and insect diversity.

Conclusions This study demonstrates a decrease in diversity in forest- agriculture landscapes depending on crop type, with lower diversity in landscapes with transgenic maize fields. Moreover, the influence of crop type on nearby forest fragments pointing to direct threats to biodiversity, and additional side effects coming from adjacent crops, as those from compounds produced by transgenic organisms. Although it is still unclear how transgenic compounds affect the different components of food webs, our results indicate that a transgenic crop negatively affects the structure of dung beetle assemblages, especially when compared to creole crops, leading to a loss of species and individuals. Therefore, even though forest fragments have helped to maintain regional diversity, changes in agricultural systems seem to affect the conservation of a group of insects with key ecosystem roles. In addition, we suggest that growing creole maize, which unfortunately is increasingly rare, is associated with communities with more species and more abundant dung beetles, promoting the conservation of these insects and their ecosystem functions.

Acknowledgments We would like to thank Oestebio cooperative for help in choosing the areas. We are grateful to the farmers for collection permission and Anderson Munarini, Gleico Mittmann, Maicon Perdersseti, Moisés Vacega and Maristela Carpintero for help during the fieldwork.VMA thanks CAPES (Ministry of Education of Brazil) for a PhD Grant and MIMH thanks CNPq (Science and Technology Ministry of Brazil, Proc. 309030/2013-7) for a Productivity Grant. This project was supported by CNPq (Call 048/13).

References Abreu, L., Cansi, E., Juriatti, C., 2007. Avaliação do rendimento sócio- econômico de variedades crioulas e híbridos comerciais de milho na microregião de Chapecó. Rev. Bras. de Agroecologia 2, 1230- 1233. Alves, V.M., Hernández, M.I.M., 2017. Morphometric modifications in Canthon quinquemaculatus Castelnau 1840 (Coleoptera: Scarabaeinae): sublethal effects of transgenic maize? Insects 8, 115. doi: 10.3390/insects8040115 56

Araújo, P.M., Nass, L.L., 2002. Caracterização e avaliação de populações de milho crioulo. Sci. Agri. 59, 589-593. doi: 10.1590/S0103-90162002000300027 Brooks, T.M., Mittermeier, R.A., Mittermeier, C.G., Fonseca, G.A.B., Rylands, A.B., Konstant, W.R., Flick, R., Pilgrim, J., Oldfield, S., Magin, G., Hilton-taylor, C.H., 2002. Habitat loss and extinction in the hotspots. Conserv. Biol. 16, 909-923. doi: 10.1046/j.1523- 1739.2002.00530.x Campos, R.C., Hernández, M.I.M., 2015a. The importance of maize management on dung beetle assemblages in Atlantic Forest fragments. PLoS ONE 10, e0145000. doi: 10.1371/journal.pone.0145000 Campos, R.C., Hernández, M.I.M., 2015b. Changes in the dynamics of functional groups in assemblages of dung beetles in Atlantic forest fragments adjacent to transgenic maize crops. Ecol. Indic. 49, 216- 227. doi: 10.1016/j.ecolind.2014.09.043 Carpentieri-Pípolo, V., Souza, A., Silva, D.A., Barreto, T.P., Garbuglio, D.D., Ferreira, J.M., 2010. Avaliação de cultivares de milho crioulo em sistema de baixo nível tecnológico. Acta Sci. Agron. 32, 229-233. doi: 10.4025/actasciagron.v32i2.430 Chapin, F.S., Zavaleta, E.S., Eviner, V.T., Naylor, R.L., Viousek, P.M., Reynolds, H.L., Hooper, D.U., Lavorel, S., Sala, O.E., Hobbie, S.E., Mack, M.C., Díaz, S., 2000. Consequences of changing biodiversity. Nature 405, 234-242. doi: 10.1038/35012241 Colwell, R.K., 2013. EstimateS, Statistical estimation of species richness and shared species from samples. Version 9. http://viceroy.eeb.uconn.edu/EstimateS/ (acessed 20 June 2015) Dutton, A., Klein, H., Romeis, J., Bigler, F., 2002. Uptake of Bt-toxin by herbivores feeding on transgenic maize and consequences for the predator Chrysoperla carnea. Ecol. Entomol. 27, 441-447. doi: 10.1046/j.1365-2311.2002.00436.x Estrada, A., Coates-Estrada, R., Dadda, A.A., Cammarano, P., 1998. Dung and carrion beetles in tropical rain forest fragments and agricultural habitats at Los Tuxtlas, Mexico. J. Trop. Ecol. 14, 577- 593. doi: 10.1017/S0266467498000418 Farias, P.M., Hernández, M.I.M., 2016. Dung beetles associated with agroecosystems of Southern Brazil: relationship with soil properties. Rev. Bras. Ciênc. Solo. 41, e0160248. doi: 10.1590/18069657rbcs20160248. Ferreira, P.I., Gomes, J.P., Batista, F., Bernardi, A.P., Costa, N.C.F., Bortoluzzi, R.L.C., Mantovani, A., 2013. Espécies potenciais para 57

recuperação de áreas de preservação permanente no planalto catarinense. Floresta Ambient. 20, 173-182. doi: 10.4322/floram.2013.003 Gardner, T.A., Barlow, J., Araujo, I.S., Ávila-Pires, T.C., Bonaldo, A.B., Costa, J.E., Esposito, M.C., Ferreira, L.V., Hawes, J., Hernández, M.I.M., Hoogmooed, M.S., Leite, R.N., Lo-Man- Hung, N.F., Malcolm, J.R., Martins, M.B., Mestre, L.A.M., Miranda-Santos, R., Overal, W.L., Parry, L., Peters, S.L., Ribeiro- Junior, M.A., da Silva, M.N.F., da Silva, C.M., Peres, C.A., 2008a. The cost-effectiveness of biodiversity surveys in tropical forests. Ecol. Lett. 11, 139–150. doi: 10.1111/j.1461-0248.2007.01133.x Gardner, T.A., Hernández, M.I.M., Barlow, J., Peres, C.A., 2008b. Understanding the biodiversity consequences of habitat change: the value of secondary and plantation forests for neotropical dung beetles. J. Appl. Ecol. 45, 883-893. doi: 10.1111/j.1365- 2664.2008.01454.x Gassmann, A.J., Petzold-Maxwell, J.L., Keweshan, R.S., Dunbar, M.W., 2011. Field-evolved resistance to Bt maize by western corn rootworm. PLoS ONE 6, e22629. doi: 10.1371/journal.pone.0022629 Gassmann, A.J., Petzold-Maxwell, J.L., Clifton, E.H., Dunbar, M.W., Hoffmann, A.M., Ingber, D.A., Keweshan, R.S., 2014. Field- evolved resistance by western corn rootworm to multiple Bacillus thuringiensis toxins in transgenic maize. Proc. Natl. Acad. Sci. USA. 111, 5141-5146. doi: 10.1073/pnas.1317179111 Halffter, G., Matthews, E.G. 1966. The natural history of dung beetles of the subfamily Scarabaeinae (Coleoptera: Scarabaeidae). Folia Entomol. Mexi. 12, 1-312. Halffter, G., Edmonds, W.D., 1982. The nesting behavior of dung beetles (Scarabaeinae): An ecological and evolutive approach, México. Halffter, G., Favila, M., 1993. The Scarabaeinae (Insecta: Coleoptera) an animal group for analysing, inventorying and monitoring biodiversity in tropical rain forest and modified landscapes. Biol. Int. 27, 15-21. Horgan, F.G., 2008. Dung beetle assemblages in forests and pastures of El Salvador: a functional comparison. Biodivers. Conserv. 17, 2961–2978. doi: 10.1007/s10531-008-9408-2 ISAAA, 2016. Global Status of Commercialized Biotech/GM Crops: 2016. ISAAA Brief No. 52. ISAAA: Ithaca, NY. http://www.isaaa.org (accessed 20 August 2017). 58

Jump, A.S., Peñuelas, J., 2006 Genetic effects of chronic habitat fragmentation in a wind-pollinated tree. Proc. Natl. Acad. Sci. USA. 103, 8096-8100. doi: 10.1073/pnas.0510127103 Leibold, M.A., Holyoak, M., Mouquet, N., Amarasekare, P., Chase, J.M., Hoopes, M.F., Holt, R.D., Shurin, J.B., Law, R., Tilman, D., Loreau, M., Gonzalez, A., 2004. The metacommunity concept: a framework for multi-scale community ecology. Ecol. Lett. 7, 601– 613. doi: 10.1111/j.1461-0248.2004.00608.x Lobo, J.M., Martin-Piera, F., Veiga C.M., 1988. Las trampas pitfall con cebo, sus posibilidades en el estudio de las comunidades coprófagas de Scarabaeiodea (Col.). I. Características determinantes de su capacidad de captura. Rev. Ecol. Biol. Sol. 25, 77-100. Losey, J.E., Rayor, L.S., Carter, M.E., 1999 Transgenic pollen harms monarch larvae. Nature. 399, 214. doi: 10.1038/20338 Marvier, M., McCreedy, C., Regetz, J., Kareiva, P., 2007. A meta- analysis of effects of Bt cotton and maize on nontarget invertebrates. Science. 316, 1475–1477. doi: 10.1126/science.1139208 Moore, R.P., Robinson, W.D., Lovette, I.J., Robinson, T.R., 2008. Experimental evidence for extreme dispersal limitation in tropical forest birds. Ecol. Lett. 11, 960-968. doi: 10.1111/j.1461- 0248.2008.01196.x Nichols, E., Larsen, T., Spector, S., Davis, A.L., Escobar, F., Favila, M. & Vulinec, K. (2007) Global dung beetle response to tropical forest modification and fragmentation: A quantitative literature review and meta-analysis. Biol. Conserv. 137, 1-19. doi: 10.1016/j.biocon.2007.01.023 Niemi, G.J., Mcdonald, M.E., 2004. Application of ecological indicators. Annu. Rev. Ecol. Evol. Syst. 35, 89 –111. doi: 10.1146/annurev.ecolsys.35.112202.130132 Palmquist, K., Salatas, J., Fairbrother, A., 2012. Pyrethroid Insecticides: Use, Environmental Fate, and Ecotoxicology, in: Perveen, F. (Ed.), Insecticides - Advances in Integrated Pest Management. Intech, Rijeka, pp. 251-278. R Core Team, 2015. R: A language and environment for statistical computing. R Foundation for Statistical Computing. http://www.R- project.org/ (accessed 02 May 2015). Ranum, P., Peña-Rosas, J.P., Garcia-Casal, M.N., 2014. Global maize production, utilization, and consumption. Ann. N. Y. Acad. Sci. 1312, 105-112. doi: 10.1111/nyas.12396 59

Scholtz, C.H., Davis, A.L.V., Kryger, U. 2009. Evolutionary Biology and Conservation of Dung Beetles, Pensoft Publishers, Sofia. Scipioni, M.C., Galvão, F., Longhi, S.J., 2013. Composição florística e estratégias de dispersão e regeneração de grupos florísticos em florestas estacionais deciduais no Rio Grande do Sul. Floresta. 43, 241-254. doi: 10.5380/rf.v43i2.27098 Silva, J.S., Ranieri, V.E.L., 2014. O mecanismo de compensação de reserva legal e suas implicações econômicas e ambientais. Ambient. Soc. 17, 115-132. Slade, E.M., Mann, D.J., Villanueva, J.F., Lewis, O.T., 2007. Experimental evidence for the effects of dung beetle functional group richness and composition on ecosystem function in a tropical forest. J. Anim. Ecol. 76, 1094-1104. doi: 10.1111/j.1365- 2656.2007.01296.x Sluyter, A., Dominguez, G., 2006. Early maize (Zea mays L.) cultivation in Mexico: Dating sedimentary pollen records and its implications. Proc. Natl. Acad. Sci. USA. 103, 1147-1151. doi: 10.1073/pnas.0510473103 Tago, D., Andersson, H., Treich, N., 2014. Pesticides and health: A review of evidence on health effects, valuation of risks, and benefit‐cost analysis. Adv. Health Econ. Health Serv. Res. 24, 203- 295. Thomé, V.M.R., Zampieri, S., Braga, H.J., Pandolfo, C., Silva, V.P., Bacic, I., Laus, J., Soldateli, D., Gebler, E., Ore, J.D., Echeverria, L., Mattos, M., Suski, P.P., 1999. Zoneamento agroecológico e socioeconômico de santa Catarina. Epagri: Florianópolis. http://ciram.epagri.sc.gov.br (accessed 13 March 2015). Tilman, D., Fargione, J., Wolff, B, D‟Antonio, C., Dobson, A., Howarth, R., Schindler, D., Schlesinger, W.H., Simberloff, D., Swackhamer, D., 2001. Forecasting agriculturally driven global environmental change. Science. 292, 281-284. doi: 10.1126/science.1057544 Turner, I.M., 1996. Species loss in fragments of tropical rain forest: a review of the evidence. J. Appl. Ecol. 33, 200-209. doi: 10.2307/2404743 Vandermeer, J.H., 2011. The ecology of agroecosystems, Jones and Bartlett Publishers, Sudbury. Vibrans, A.C., McRoberts, R.E., Moser, P. & Nicoletti, A.L., 2013 Using satellite image-based maps and ground inventory data to estimate the remaining Atlantic forest in the Brazilian state of 60

Santa Catarina. Remote Sens. Environ. 130, 87-95. doi: 10.1016/j.rse.2012.10.023 Wang, Y., Naumann, U., Wright, S.T. & Warton, D.I., 2012. Mvabund: an R package for model-based analysis of multivariate abundance data. Methods Ecol. Evol. 3, 471-474. doi: 10.1111/j.2041- 210X.2012.00190.

Supporting Information

Agricultural Area Herbicide Insectice Fungicide input use CO1 1 0 0 1 CO2 1 0 0 1 CO3 1 0 0 1 CO4 1 1 0 2 CO5 1 1 0 2 CR1 0 0 0 0 CR2 0 0 0 0 CR3 0 0 0 0 CR4 1 0 0 1 CR5 0 0 0 0 TR1 1 1 1 3 TR2 1 1 1 3 TR3 1 1 0 2 TR4 1 0 0 1 TR5 1 0 0 1

Table S2. Tree species found in 15 Atlantic Forest fragments nearby maize fields in the region of São Miguel do Oeste, state of Santa Catarina, Brazil. The number of trees of each species found in landscapes with distinct types of maize are indicated (CO: conventional, CR: creole and TR: transgenic maize). Tree functional groups (FG) area as follow: pioneer (P), secondary light-demanding (SL), and secondary shade tolerant tree species (SS).

Species CO CR TR FG Achatocarpus praecox 0 1 0 SL Albizia niopoides 12 0 1 SL Alchornea sidifolia 0 0 5 SL Allophylus edulis 4 4 5 SL Aloysia virgata 0 1 0 SL Annona neosalicifolia 1 0 0 SL Annona sylvatica 1 1 1 SL Araucaria angustifolia 1 0 0 SL Aspidosperma australe 1 1 0 SS Balfourodendron 2 3 1 SL riedelianum Banara tomentosa 0 1 0 SS Bauhinia forficata 6 0 1 SL Boehmeria caudata 1 0 0 P Cabralea canjerana 1 0 1 SL Calliandra foliolosa 1 0 0 SL Campomanesia 0 1 3 SS xanthocarpa Casearia decandra 1 0 1 SS Casearia sylvestris 1 2 0 SL Celtis iguanae 1 0 0 P Chrysophyllum 2 3 0 SS gonocarpum Chrysophyllum 1 0 6 SL marginatum Citrus reticulata 1 0 3 SL Citrus xaurantium 1 0 0 SL Cordia americana 0 2 5 P Cordia trichotoma 2 14 2 P Cordyline spectabilis 0 0 1 SL Cupania vernalis 5 1 1 SL Diatenopteryx sorbifolia 1 3 9 SL Erythrina falcata 0 2 0 SL Eucalyptus grandis 0 1 0 P Eugenia involucrata 1 0 0 SL Eugenia uniflora 0 1 1 SL Ficus luschnathiana 0 1 0 SL 63

Gymnanthes concolor 0 3 0 SS Helietta apiculata 0 0 15 P Heliocarpus americanus 6 4 0 P Holocalyx balansae 0 1 0 SL Hovenia dulcis 2 0 5 P Inga marginata 3 1 0 SL Jacaranda micrantha 0 1 1 P Lonchocarpus campestris 21 4 2 SL Lonchocarpus 0 0 2 SL muehlbergianus Luehea divaricata 0 18 1 SL Machaerium 1 1 0 SL paraguariense Machaerium stipitatum 11 12 4 SL Maclura tinctoria 0 0 1 SL Matayba elaeagnoides 7 2 1 SL Morus nigra 1 0 1 P Myrocarpus frondosus 6 6 7 SL Myrsine coriacea 0 0 2 P Myrsine umbellata 1 1 0 SL Nectandra lanceolata 1 0 0 SL Nectandra 9 5 8 SL megapotamica Ocotea diospyrifolia 0 1 0 SL Ocotea puberula 1 1 6 SL Ocotea pulchella 0 1 0 SL Parapiptadenia rigida 3 29 32 SL Phytolacca dioica 4 0 0 SL Pilocarpus pennatifolius 3 1 2 SS Pinus eliotti 14 0 0 P Prunus myrtifolia 0 2 1 SL Rauvolfia sellowii 2 0 0 SL Ruprechtia laxiflora 1 0 5 SL Sambucus australis 3 1 0 P Schefflera morototoni 0 1 0 SL Sebastiania brasiliensis 0 5 3 SL Sebastiania 1 2 9 P commersoniana Seguieria guaranitica 0 3 1 SL Solanum mauritianum 2 0 0 P Solanum pseudoquina 2 0 0 SL Sorocea bonplandii 2 0 1 SS Strychnos brasiliensis 2 1 7 SL Styrax leprosus 0 0 1 SL Trichilia catigua 0 0 1 SS 64

Trichilia claussenii 0 3 0 SS Trichilia elegans 0 0 1 SS Urera baccifera 7 2 0 SL Vassobia breviflora 0 1 0 SL Vitex megapotamica 1 1 0 SL Zanthoxylum petiolare 1 0 1 P Zanthoxylum rhoifolium 1 0 0 P Total 167 157 168

Fig. S1 Landscapes of the blocks with two landscape units (cultivation and forest remnants) of conventional corn areas in the region of São Miguel do Oeste, state of Santa Catarina, Brazil

Fig. S2 Landscapes of the blocks with two landscape units (cultivation and forest remnants) of creole corn areas in the region of São Miguel do Oeste, state of Santa Catarina, Brazil

Fig. S3 Landscapes of the blocks with two landscape units (cultivation and forest remnants) of transgenic corn areas in the region of São Miguel do Oeste, state of Santa Catarina, Brazil

Fig. S4 Scheme of blocks consisting of two landscape units: maize fields and nearby forest fragments. Blocks were replicated five times for each type of maize grown in the landscape (conventional, creole and transgenic), totaling 15 blocks and 30 landscape units where dung beetles were sampled in the region of São Miguel do Oeste, state of Santa Catarina, Brazil.

Fig. S5 Representation of the pitfall traps arrangement to dung beetle collection in the region of São Miguel do Oeste, SC, Brasil. Each area had 10 traps: five baited with human feces and five with pork carrion. The traps with same bait were placed 10 meters from each other and five meters with different baits. The sampling unit was the pair of traps.

Fig. S6 Residuals resulting from the fitted multivariate generalized linear model (GLMmv).

ARTIGO II: MORPHOMETRIC MODIFICATIONS IN CANTHON QUINQUEMACULATUS CASTELNAU 1840 (COLEOPTERA: SCARABAEINAE): SUBLETHAL EFFECTS OF TRANSGENIC MAIZE?

Victor Michelon Alves1, Malva Isabel Medina Hernández1

1 Programa de Pós-Graduação em Ecologia, Departamento de Ecologia e Zoologia, Universidade Federal de Santa Catarina, Florianópolis, Santa Catarina, 88040–900

Artigo publicado no periódico Insects. Outubro 2017. DOI: 10.3390/insects8040115

Morphometric modifications in Canthon quinquemaculatus Castelnau 1840 (Coleoptera: Scarabaeinae): sublethal effects of transgenic maize?

Abstract The effects of transgenic compounds on non-target organisms remain poorly understood, especially in native insect species. Morphological changes (e.g. changes in body size and shape) may reflect possible responses to environmental stressors, like transgenic toxins. The dung beetle Canthon quinquemaculatus (Coleoptera: Scarabaeinae) is a non- target species found in transgenic crops. We evaluated whether C. quinquemaculatus individuals inhabiting corn fields cultivated with different seed types (conventional, creole and transgenic) present modifications in body shape compared to individuals inhabiting adjacent native forest fragments. We collected C. quinquemaculatus specimens across an agricultural landscape in southern Brazil, during the summer of 2015. Six populations were sampled: three maize crop populations each under a different seed type, and three populations of adjacent forests. After sampling, specimens were subjected to morphometric analyses to discover differences in body shape. We chose fifteen landmarks to describe body shape, and morphometric data were tested with Procrustes ANOVA and Discriminant Analysis. We found that body shape did not differ between individuals collected in conventional and creole crops with their respective adjacent forests (p > 0.05); however, transgenic crop populations differed significantly from those collected in adjacent forests (p < 0.05). Insects in transgenic maize are more oval and have a retraction in the abdominal region, compared with the respective adjacent forest, this result shows the possible effect of transgenic crops on non-target species. This may have implications for the ecosystem service of organic matter removal, carried out by these organisms.

Keywords: agriculture; body shape; dung beetle; ecology; morphology

Introduction Body shape is directly associated with several important ecological aspects, such as high-speed displacement capabilities, and consequently predator–prey interactions [1,2]. Shape characteristics can give advantages to males in sexual selection, having greater chances of mating [3]. Modifications in body shape offer advantages to certain types of behavior, such as foraging [4,5], reproduction [6] and habitat use [7]. Body morphology is also linked with community structuring since differences in shape are usually associated with different feeding and nesting strategies. Consequently, shape can be associated with better adapted competitors with improved foraging strategies in a particular community [8]. In addition to providing important information on species ecology, body shape may also indicate the effects of possible environmental stressors on individuals. These environmental stressors may be natural, such as ultraviolet radiation [9], or may be anthropic, like pollution [10]. Morphological alterations may imply possible ecological consequences, such as reduction in feed and displacement capacities [11,12]. Thus, in the face of environmental disturbances, body shape alterations may be classified as sublethal effects, and can affect an individual‟s fitness [9,13]. Sublethal effects are defined as the physiological or behavioral changes in individuals exposed to chemical compounds [14]. These changes reduce individual fitness and have already been described in various non-target insects such as butterflies [15], bees [16] and dung beetles [17]. Sublethal effects of chemical compounds associated with agricultural practices have been observed in dung beetles. Dung beetles fed with feces containing Ivermectin (a chemical product used to control parasites in cattle) presented behavioral changes, increased developmental time, reduced fecundity and oviposition, and reduced cephalic capsule size [18,19,20]. Behavioral and physiological changes were also described in dung beetles fed with cattle feces containing transgenic maize. When fed with transgenic products, dung beetles showed lower dispersion velocity and alterations in the immune system [21]. These sublethal effects decrease the individual‟s fitness, leading to abnormalities in their development, behavior and morphology. Fitness reduction may be evidenced by changes in individual size [22] or shape, particularly in holometabolous insects where the adult stage is greatly influenced by larval development. Therefore, larvae that develop in poor environments, with few resources, or under some kind of specific crop 75

system (e.g., transgenic crops or under insecticidal and fungicidal effect) may result in adults presenting changes in shape and size. Maize (Zea mays L.) is the second largest crop production in the world [23]. Maize production is associated, among other factors, with seed differences. The use of genetically modified seeds is growing globally, reaching approximately 180 million hectares in 2014 [24]. Since the beginning of its implementation in the mid-1980s, questions about their environmental safety have been raised [25]. Some studies have described the effectiveness and apparent safety of transgenic maize in reducing economically important insect populations and not harming non-target organisms [26,27]. On the other hand, several studies have shown target insect resistance development [28], accumulation of transgenic toxin compounds in the food chain [29], and lethal effects on non-target organisms [30]. Conventional seed varieties are severely affected by insects of economic importance, since the combination of food abundance with lack of natural enemies (due to native vegetation suppression) favors overpopulation of phytophagous insects, mainly Lepidoptera larvae. Creole seeds (landraces) normally are less productive than other seeds; however, they have increased genetic variability, and more resistance to phytophagous insects and specific environmental conditions [31]. These landraces can be productive even in unfavorable conditions [32], and represent a genetic heritage since they are selected by the farmers themselves, through the selection of more adapted individuals [33]. Canthon quinquemaculatus Castelnau 1840 is a non-target species that feeds on decaying material, and thus is engaged in nutrient cycling, promoting environmental services regulating physico-chemical soil properties [34,35]. C. quinquemaculatus is a widespread species throughout South America and it feeds on feces and carcasses of mammals and birds. These animals may consume different types of maize. Bt toxin can be transferred through the trophic chain: individuals of Orius majusculus (an insect predator) contained Bt toxins when feeding on spider mites reared on transgenic maize [36]. Thus, through the trophic chain these Bt toxins can accumulate and reach C. quinquemaculatus populations. In addition to the accumulation of transgenic toxins, individuals may suffer from different types of inputs, such as insecticides, which are used in maize crops. These inputs are greater in transgenic and conventional seeds, and are extremely low in creole seeds. Under the assumption that changes in individual shape indicate possible effects on non-target species physiology, the aim of this study was to evaluate whether C. quinquemaculatus individuals 76

present morphometric differences in populations found within different maize crops (conventional, creole, and transgenic seeds) when compared with populations found in adjacent control populations inhabiting forest remnants.

Materials and Methods Population sampling occurred in the region of São Miguel do Oeste, Santa Catarina State, Brazil (26°43′31″ S, 53°31′05″ W). The region is composed of a mosaic of Atlantic forest fragments and maize crops. Local climate is humid subtropical (Cfa) according to the Köppen classification, with an annual average temperature between 16.3 °C and 17.9 °C, and rainfall between 1790 mm and 2280 mm [37]. Samples were collected during the summer of 2015 in maize crops with different profiles: five areas of conventional maize, five areas of creole maize, and five areas of transgenic maize. Furthermore, 15 adjacent forest remnants, one for each crop area, were sampled in order to have an independently replicated control population for each type of crop. The insects were collected using ten pitfall traps per area, five baited with human feces (10 g) and five with pig carrion (10 g) during 48 h. The traps were spaced 10 m apart within the same area, and a minimum distance of 50 m was established between crop traps and forest fragments to avoid interference (Appendix A Figure A1). Captured insects were fixed in 70% alcohol and deposited in the entomological collection of the Federal University of Santa Catarina (UFSC). From a total of 231 specimens of C. quinquemaculatus collected, we used 157 individuals for morphometric analyses. We tried to balance the proportion between males and females in the most equitable way possible. Thus, we measured 12 specimens from conventional maize crops (6♀, 6♂); 17 from creole crops (11♀, 6♂); 28 from transgenic crops (14♀, 14♂); 36 from forest fragments adjacent to conventional crops (18♀, 18♂); 30 from forest fragments adjacent to creole crops (10♀, 20♂); and 34 from forests adjacent to transgenic crops (17♀, 17♂). We defined the body shape through the placement of 15 anatomical landmarks. These 15 landmarks were chosen because they capture all morphological variations describing body shape in both anteroposterior and ventral dorsal axis, and these landmarks have already been successfully used to describe dung beetle body shape [8]. Each landmark corresponded to a point in space defined by three- dimensional cartesian coordinates (x, y, z) [38] as follows: (1) anterior margin of the head; (2) eye position; (3) division between the pronotum 77

and the elytra; (4) division between the thorax and the abdomen; (5) posterior margin of the abdomen; (9) point of insertion of the anterior legs; (10 and 11) points of insertion of the central legs; (12) point of insertion of the posterior legs; (13) anterior point of convergence between elytra; (14) central point (mid-line) of convergence between elytra; and 15) posterior margin (along the mid-line) of the elytra. Points 6, 7 and 8 corresponded to points 4, 3 and 2 for the other side of the body, respectively [8]. Insects were photographed using Canon T3i camera in dorsal, ventral and lateral view; therefore, body shape was captured in three dimensions (Figure 1). A standard protocol for photographing specimens was established: insects were photographed on millimeter paper on a fixed surface to minimize the position effects. The camera was attached to a tripod and kept at a distance of 10 cm from each specimen at a perpendicular angle. Each individual had a tag, and for digitalization of the landmarks the individuals were mixed, so the marking was blind.

Figure 1. Landmarks (red dots) used in shape analysis of C. quinquemaculatus. (A) Lateral view; (B) Graphical representation of body shape based on 15 landmarks, adapted from Hernández et al. [8]. Landmarks in parentheses correspond to the same region on the other side of the body.

Coordinates from each landmark underwent Generalized Procrustes Analysis (GPA). The GPA is a three-step multivariate technique that assesses body shape eliminating the effects of size, position and orientation. The first step of the GPA is to center each landmark at the cartesian origin, eliminating the position effect. Afterwards, the landmark coordinates are scaled, eliminating the effect of size. Lastly, the landmark coordinates are rotated around the origin, which removes the orientation effect. Through these three steps, size and orientation data are removed leaving only body shape information [39]. We performed a canonical variable analysis (CVA) to view each individual‟s position in the multivariate space. The overall shape difference was primarily tested with a Procrustes ANOVA, and to test the differences between maize crop populations and adjacent forest populations we performed a discriminant analysis (DA) based on the Mahalanobis distance. The reliability of a population‟s individuals was determined by cross-validation to calculate the correct percentage of classification of the individuals in the original population, based only on the morphometric variation. Anatomical landmarks were inserted into specimens with the softwares TpsDig 2.16 [40] and tpsUtil1.53 [41]. The GPA, CVA and DA were performed in Morpho J 1.05d software [42].

Results The shape of the individuals was assessed from the 15 landmarks and each specimen was represented by a point in Figure 2. Together, the first (CV1) and second (CV2) canonical variables in the CVA explained 70% of the observed body shape variation. The CVA showed that maize crop populations have mostly positive values in CV1 (red, black and green points in the Figure 2). On the other hand, forest fragment populations were mostly negative (light blue, blue and pink) and inferred a shape distinction between the insects that inhabit maize crops and those from the adjacent forests.

Figure 2. Position of C. quinquemaculatus populations in morphometric space formed by two main axes of Canonical Variable Analysis, collected in the region of São Miguel do Oeste, SC, Brazil. Conventional (Co), Creole (Cr) and Transgenic (Tr); ellipses indicate a 95% confidence interval.

The CV1 (which accounted for 52% of the variance) organized individuals in a flat (negative scores) to oval (positive scores) body shape gradient; forest insects corresponded to a flat body shape and maize crop insects to an oval body shape (see the population‟s shape design in both corners of Figure 2). Populations with positive CV1 (maize crop populations) (body shape in the right corner of Figure 2) had the landmarks of the dorsal region displaced above average shape, and landmark 5 (corresponding to abdominal region) displaced towards the body center. Together, these landmark displacements indicated a pronounced oval shape and an abdominal contraction (Figure 2 right corner). Populations with negative scores in CV1 (forest fragment populations) had landmark 5 (abdominal region) dislocated away from body center, and had the landmarks of the dorsal region displaced toward the body center. These rearrangements resulted in a flattened body shape (Figure 2, left corner). Regarding CV2, there was no clear distinction between insects, insects with positive scores had dorsal region landmarks displaced 80

towards the body center, and the ventral region landmarks dislocated away from the body center. These landmark displacements indicated a retraction on the anterior–posterior axis of body (Figure 2, designs at the top of CV2). Insects with negative scores had the dorsal region landmarks displaced away from the body center, and the ventral region landmarks dislocated towards the body center. This landmark configuration indicated an extension of the body‟s anterior–posterior axis (Figure 2, designs at bottom of CV2). The cross-validation tests indicated that an average of 75% (ranging between 58% and 90%) of individuals were correctly allocated to their respective populations. Maize crop populations had a high validation (ranging between 64% and 90%), including transgenic maize crop populations with 82% (Table 1).

Table 1. Percentage of correct classification of C. quinquemaculatus populations based on body morphology and discriminant analysis (DA). Population Allocation value (%) Conventional – Maize crop 90.66 Creole – Maize crop 64.70 Transgenic - Maize crop 82.14 Conventional - Adjacent forest 58.33 Creole – Adjacent forest 73.33 Transgenic – Adjacent forest 82.85 Overall classification accuracy 75.33

While analyzing all the populations we observed that they were different in relation to body shape (Procrustes ANOVA, F = 2.28, Pillai‟s Trace = 2.03, p < 0.0001). We observed that the shape of the populations of creole and conventional crops were not significantly different from their respective adjacent forest areas (DA, p > 0.05). However, populations from transgenic maize crops differed in body shape from the populations of adjacent forests (DA, p = 0.0002) (Figure 3). The transgenic maize crop populations have the dorsal region landmarks higher up than the average shape of insects from adjacent forests (landmarks 13, 14 and 15 in Figure 3), and landmark 5 shifted to the center of the body. Consequently, this configuration indicates that populations of transgenic crops had a more oval shape and an abdominal retraction compared to individuals from adjacent forest populations (Figure 3). 81

Figure 3. (A) 2D Graphical reconstruction of C. quinquemaculatus body shape. Gray lines show body shape of transgenic maize crop populations, and black lines show adjacent forest populations; (B) 3D Graphical reconstruction of C. quinquemaculatus body shape. Green dots: transgenic maize populations. Pink dots: adjacent forest populations.

Discussion We observed different body shape patterns for C. quinquemaculatus individuals living in different habitats; populations found in transgenic maize crops presented significant morphometric modifications when compared to adjacent forest fragments. To date no published studies have reported changes in the entire body shape in response to the use of transgenic organisms in an agricultural environment. Studies with Diabrotica virgifera virgifera C. (Coleoptera: Chrysomelidae) collected in different types of maize crop, including transgenic crops, showed that the head of males inhabiting in transgenic maize was significantly smaller than in other treatments [43]. Similarly, individuals of Papilio machaon L. (Lepidoptera: Papilionidae) fed with 82

transgenic pollen had lower body mass [22]. These results indicated that immature stages (larvae) could be exposed to sublethal doses of toxic transgenic maize, resulting in smaller adults. While implications of body size as a result of environmental effects are well documented [44], ecological implications of changes in body shape remain poorly known. Although we did not detect changes in the body size of C. quinquemaculatus, we found differences in adult morphometrics, leading us to consider that the larvae or adults may have suffered effects derived from feeding on transgenic toxins; furthermore, laboratory experiments are needed to test for differences in the morphology of adults fed with transgenic toxins. With respect to body size, several studies failed to detect changes in insect populations grown in transgenic crops [45,46]. Inability of body size and mass to detect transgenic effects on non-target insect species suggests that both variables are less sensitive than shape, and using body shape to detect sublethal effects may be more efficient and informative. Changes in shape might have a major genetic influence compared to changes in size and mass, which are mainly driven by environmental factors [47]. Changes in body shape or in the shape of functional traits (e.g., wings, legs) can greatly affect some ecological attributes like sexual selection [48], adaptation [49], competition [8], development [50], and predation [51], causing significant changes in important populational parameters such as density, morphological variation, fitness, evolution and survival. Very few studies have evaluated changes in body shape or in the shape of bodily structures in response to transgenic compounds. Rahman et al. [52] observed alterations to skull shape in transgenic tilapia fish (Oreochromis niloticus). Studies with Pacific Salmon (Oncorhynchus spp.) showed changes in the shape of the head, tail and abdominal regions in response to a transgenic insertion [53]. Studies with transgenic Coho Salmon (O. kisutch) showed body shape differences between transgenic individuals and wild individuals even if they developed at the same temperatures [12]. This shape difference observed in Oncorhynchus spp. can affect its ability to swim, escape from predators and migrate. Similarly, Dunham et al. [54] found body shape differences in transgenic common carp (Cyprinus carpio). Furthermore, animals fed with transgenic organisms showed morphometric alterations; rats fed with transgenic potatoes presented shape alterations in the ileum [55]. Shape alterations in genetically modified organisms have also been described for plants; for example, Armon et al. [56] found differences in the leaf shape of wild and 83

genetically modified Arabidopsis sp. specimens, suggesting that gene manipulation may interfere with growth regulating systems. We have indications that transgenic compounds accumulating in the trophic chain affect the dung beetles, causing abnormalities in their physiology and behavior [21]. This pattern may be a product of a sum of factors, like historical management or use of different kinds of insecticides. Habes et al. [57] demonstrated morphometric alterations in Blatella germanica ovaries after insecticide application. Moreover, the management associated with transgenic crops alters the structure of the dung beetle community, reducing species abundance [30]. Therefore, the results from C. quinquemaculatus have given us another clue as to how transgenic compounds may be affecting dung beetles. As observed for C. quinquemaculatus, changes in insect morphometrics may lead to a decrease in insect fitness in response to the use of transgenics.

Conclusions We have shown that the transgenic maize populations of C. quinquemaculatus are associated with changes in body shape, which may be a possible sub-lethal effect of the transgenic crop in this non- target species. These effects may have repercussions on the ecosystem functions performed by these insects, such as organic matter removal and regulation of soil physico-chemical properties.

Acknowledgments We are grateful to Paulo Emilio Lovato (Department of Rural Engineering/UFSC) for coordinating the project “Fortalecimento das condições de produção e oferta de sementes de milho para a produção orgânica e agroecológica do Sul do Brasil” (CNPq Call 048/13). VMA would like to thank CAPES (Ministry of Education of Brazil) for a PhD Grant and MIMH would like to thank CNPq (Proc. 309030/2013-7) for a Research Productivity Grant.

Author Contributions Both authors contributed equally in this work.

References

1. Burns, J.G.; di Nardo, P.; Rodd, F.H. The role of predation in variation in body shape in guppies Poecilia reticulata: a comparison of field and common garden phenotypes. J. Fish. 84

Biol. 2009, 75, 1144-1157, DOI: 10.1111/j.1095- 8649.2009.02314.x. 2. Dayton, G.H.; Saenz, D.; Baum, K.A.; Langerhans, R.B.; DeWitt, T.J. Body shape, burst speed and escape behavior of larval anurans. Oikos 2005, 111, 582-591, DOI: 10.1111/j.1600-0706.2005.14340.x. 3. Rowe, L.; Arnqvist, G. Sexual selection and the evolution of genital shape and complexity in water striders. Evolution 2012, 66, 40-54, DOI: 10.1111/j.1558-5646.2011.01411.x. 4. Walker, J.A. Ecological morphology of lacustrine threespine stickleback Gasterosteus aculeatus L. (Gastoresteidae) body shape. Biol. J. Linn. Soc. 1997, 61, 3-50, DOI: 10.1111/j.1095- 8312.1997.tb01777.x. 5. Keast, A.; Webb, D. Mouth and body form relative to feeding ecology in the fish fauna of a small lake, lake Opinicon, Ontario. J. Fish. Res. Board. Can. 1966, 23, 1845-1874, DOI: 10.1139/f66-175. 6. Forsman, A.; Shine, R. Parallel geographic variation in body shape and reproductive life history within the Australian scincid lizard Lampropholis delicata. Funct. Ecol. 1995, 9, 818-828, DOI: 10.2307/2389979. 7. Vanhooydonck, B.; Damme, R.V. Evolutionary relationships between body shape and habitat use in lacertid lizards. Evol. Ecol. Res. 1999, 1, 785-805. 8. Hernández, M.I.M.; Monteiro, L.R.; Favila, M.E. The role of body size and shape in understanding competitive interactions within a community of Neotropical dung beetles. J. Insect. Sci. 2011, 11, 1-14, DOI: 10.1673/031.011.0113. 9. Bernal, M.H.; Alton, L.A.; Cramp, R.L.; Franklin CE. Does simultaneous UV-B exposure enhance the lethal and sub-lethal effects of aquatic hypoxia on developing anuran embryos and larvae? J. Comp. Physiol. B. 2011, 181, 973-980, DOI: 10.1007/s00360-011-0581-3. 10. Primost, M.A.; Bigatti, G.; Márquez, F. Shell shape as indicator of pollution in marine gastropods affect by imposex. Mar. Freshwater. Res. 2015, 67, 1948-1954, DOI: 10.1071/MF15233. 85

11. Chan, K.Y.K.; Grünbaum, D.; O‟Donnell, M.J. Effects of ocean-acidification-induced morphological changes on larval swimming. J. Exp. Bio. 2011, 214, 3857-3867, DOI: 10.1242/jeb.054809. 12. Lõhmus, M.; Susdström, L.F.; Björklund, M.; Devlin, R.H. Genotype-temperature interaction in the regulation of development, growth, and morphometrics in wild-type, and growth-hormone transgenic coho salmon. PLoS One 2010, 5, e9980, DOI: 10.1371/journal.pone.0009980. 13. Woodley, S.K.; Matters, B.M.; Yates, E.K.; Relyea, R.A. Exposure to sublethal concentrations of a pesticide or predator cues induces changes in brain architecture in larval amphibians. Oecologia 2015, 179, 655-665, DOI: 10.1007/s00442-015- 3386-3. 14. Desneux, N.; Decourtye, A.; Delpuech, J.M. The sublethal effects of pesticides on benefical arthropods. Annu. Rev. Entomol. 2007, 52, 81-106, DOI: 10.1007/s00360-011-0581-3. 15. Pecenka, J.R.; Lundegren, J.G. Non-target effects of clothianidin on monarch butterflies. Sci Nat. 2015, 102, 1-4, DOI: 10.1007/s00114-015-1270-y. 16. Dively, G.P.; Embrey, M.S.; Kamel, A.; Hawthorne, D.J.; Pettis, J.S. Assessment of Chronic Sublethal Effects of Imidacloprid on Honey Bee Colony Health. PLoS One 2015, 10, e0118748, DOI: 10.1371/journal.pone.0118748. 17. Verdú, J.R.; Cortez, V.; Ortiz, A.J.; González-Rodríguez, E.; Martinez-Pinna, J., Lumaret, J.P.; Lobo, J.M.; Numa, C.; Sánchez-Piñero, F. Low doses of ivermectin cause sensory and locomotor disorders in dung beetles. Sci. Rep. 2015, 5, 13912, DOI: 10.1038/srep13912. 18. Sommer, C.; Nielsen, B.O. Larvae of the dung beetle Onthophagus gazella F. (Col., Scarabaeidae) exposed to lethal and sublethal ivermectin concentrations. J. Appl. Entomol. 1992, 114, 502-509, DOI: 10.1111/j.1439-0418.1992.tb01158.x 19. Rosales, M.C.; Martínez, M.I.; López-Collado, J.; Vargas- Mendoza, M.; González-Hernández, H.; Fajersson, P. Effect of ivermectin on the survival and fecundity of Euoniticellus intermedius (Coleoptera: Scarabaeidae). Rev. Biol. Trop. 2012, 60; 333-345. 86

20. Pérez-Cogollo, L.C.; Rodriguez-Vivas, R.I.; Delfín-González, H.; Reyes-Novelo, E.; Ojeda-Chi, M.M. Lethal and sublethal effects of Ivermectin on Onthophagus landolti (Coleoptera:Scarabaeidae). Environ. Entomol. 2015, 44, 1634- 1640, DOI: 10.1093/ee/nvv139. 21. Campos, RC. Influência do cultivo de milho transgênico em organismos não-alvo (Coleoptera: Scarabaeinae) e da ingestão indireta através da cadeia trófica. Doctoral thesis, Federal University of Santa Catarina, Florianópolis, July 2016. 22. Lang, A.; Vojtech, E. The effects of pollen consumption of transgenic Bt maize on the common swallowtail, Papilio machaon L. (Lepidoptera, Papilionidae). Basic. Appl. Ecol. 2006, 7, 296-306., DOI: 10.1016/j.baae.2005.10.003. 23. FAO, Food and Agriculture Organization of the United Nations. Available online: http://www.fao.org./faostat (accessed 27 September 2017). 24. ISAAA, Global status of commercialized Biotech/GM Crops: 2014. Brief 49-2014: Executive Summary. Avaliable online: http://isaaa.org (accessed 20 November 2016). 25. Ceccarelli, S. GM Crops, Organic Agriculture and Breeding for Sustainability. Sustainability 2014, 6, 4273-4286, DOI: 10.3390/su6074273. 26. Zwahlen, C.; Hilbeck, A.; Nentwig, W. Field decomposition of transgenic Bt maize residue and the impact on non-target soil invertebrates. Plant. Soil 2007, 300, 245-257, DOI: 10.1007/s11104-007-9410-6. 27. Daly, T.; Buntin, G.D. Effect of Bacillus thuringiensis transgenic corn for Lepidopteran control on nontarget arthropods. Environ. Entomol. 2005, 34, 1292-1301, DOI: 10.1093/ee/34.5.1292. 28. Gassmann, A.J.; Petzold-Maxwell, J.L.; Keweshan, R.S; Dunbar, M.W. Field-evolved resistance to Bt maize by Western Corn Rootworm. PLoS ONE 2011 6, e22629.DOI: 10.1371/journal.pone.0022629 29. Hilbeck, A.; Moar, W.J; Pusztai-Carey M.; Filipini A.; Bigler F. Toxicity of Bacillus thuringiensis Cry1Ab toxin to the predator Chrysoperla carnea (Neuroptera, Chrysopidae). Environ. Entomol. 1998, 27, 1255–1263, DOI: 10.1093/ee/27.5.1255. 87

30. Campos RC, Hernández MIM. The importance of maize management on dung beetle communities in Atlantic Forest fragments. PLoS One 2015, 10, e0145000, DOI: 10.1371/journal.pone.0145000. 31. Sandri, C.A.; Tofanelli, M.B.D. Milho crioulo: uma alternativa para retabilidade no campo. Pesq Agropec. Trop. 2008, 38, 59- 61. 32. Spaner, D.; Brathwaite, R.A.I.; Mather, D.E. Comparison of open-pollinated stress-tolerant and landrace maize for production under stress conditions in Trinidad. Maydica 1995, 40, 331-337. 33. Ferreira, J.M.; Moreira, R.M.P.; Hidalgo, J.A.F. Capacidade combinatória e heterose em populaçoes de milho crioulo. Cienc. Rural 2008, 39, 332-339, DOI: 10.1590/S0103- 84782008005000058. 34. Halffter, G.; Matthews, E.G. The natural history of dung beetles of the subfamily Scarabaeinae (Coleoptera: Scarabaeidae). Folia Entomol. Mex. 1966, 12-14, 1-312. 35. Cantil, L.F.; Sánchez, M.V.; Dinghi, P.A.; Genise, J.F. Food relocation behavior, nests, and brood balls of Canthon quinquemaculatus Laporte de Castelnau (Coleoptera: Scarabaeidae: Scarabaeinae). Coleopts. Bull. 2014, 68, 199-208, DOI: 10.1649/0010-065X-68.2.199. 36. Obrist, L.B.; Dutton, A.; Albajes, R.; Bigler, F. Exposure of arthropod predators to Cry1Ab toxin in Bt maize fields. Ecol. Entomol. 2006, 31, 143-154, DOI: 10.1111/j.0307- 6946.2006.00762.x. 37. Thomé, V.M.R.; Zampieri, S.; Braga, H.J.; Pandolfo, C.; Silva, V.P.; Bacic, I.; Laus, J.; Soldateli, D.; Gebler, E.; Ore, J.D; Echeverria, L.; Mattos, M.; Suski, P.P. Zoneamento agroecológico e socioeconômico de santa Catarina. Epagri: Florianópolis. Available online: http://ciram.epagri.sc.gov.br (Accessed 03 May 2016). 38. Bookstein, F.L. Morphometric tools for landmark data: Geometry and Biology, 1st ed.; Cambridge University Press: Cambridge, England, 1991; pp. 1-455, ISBN: 0521585988. 88

39. Rohlf, F.J.; Slice, D. Extensions of the Procrustes method for optimal superimposition of landmarks. Syst. Zool. 1990, 39, 40- 59, DOI: 10.2307/2992207. 40. Rohlf FJ. TpsDig, Version 2.16. Department of Ecology and Evolution, State University of New York at Stony Brook. 2010. Available online: http://life.bio.sunysb.edu/morph/bibr28 (accessed 20 May 2016). 41. Rohlf FJ. TpsUtil, File Utility Program. Version 1.53. Department of Ecology and Evolution, State University of New York at Stony Brook. 2012. Available online: http://life.bio.sunysb.edu/morph (accessed 25 May 2016). 42. Klingenberg, C.P. MorphoJ: an integrated software package for geometric morphometrics. Mol. Ecol. Resour. 2011, 11, 353- 357, DOI: 10.1111/j.1755-0998.2010.02924.x. 43. Murphy, A.F.; Seiter, N.J.; Krupke, C.H. The impact of Bt maize as a natal host on adult head capsule width in field populations of western corn rootworm. Entomol. Exp. Appl. 2011, 139, 8-16, DOI: 10.1111/j.1570-7458.2011.01100.x. 44. Peters, R.H. The ecological implications of body size, 1st ed.; Cambridge University Press: Cambridge, England, 1983, pp. 1- 344, ISBN: 052128886X. 45. Al-Deeb, M.A.; Wilde G.E. Effect of bt corn expressing the Cry3Bb1 toxin on western corn rootworm (Coleoptera: Chrysomelidae) Biology. J. Kansas. Entomol. Soc. 2005, 78, 142-152. DOI: 10.2317/0403.04.1 46. Lundgren, J.G.; Wiedenmann, R.N. Coleopteran-specific Cry3Bb toxin from transgenic corn pollen does not affect the fitness of a nontarget Species, Coleomegilla maculata DeGeer (Coleoptera: Coccinellidae). Environ. Entomol. 2002, 31, 1213- 1218, DOI: 10.1603/0046-225X-31.6.1213. 47. Cavicchi. S.; Giorgi, G.; Mochi, M. Investigation on early divergence between populations of Drosophila melanogaster kept at different temperatures. Genetica 1978, 48, 81-87. 48. Marsteller, S.; Adams, D.C.; Collyer, M.L.; Condon, M. Six cryptic species on a single species of host plant: morphometric evidence for possible reproductive character displacement. Ecol. Entomol. 2009, 34, 66-73, DOI: 10.1111/j.1365- 2311.2008.01047.x. 89

49. Alves, V.M.; Moura, M.O.; de Carvalho, C.J.B. Wing shape is influenced by environmental variability in Polietina orbitalis (Stein) (Diptera:Muscidae). Rev. Bras. entomol. 2016, 60, 150- 156, DOI: 10.1016/j.rbe.2016.02.003 50. Winans, G.A.; Nishioka, R.S. A multivariate description of change in body shape of coho salmon (Oncorhynchus kisutch) during smoltification. Aquaculture 1987, 66, 235-245, DOI: 10.1016/0044-8486(87)90109-8. 51. McCollum, S.A.; Leimberger, J.D. Predator-induced morphological changes in an amphibian: predation by dragonflies affects tadpole shape and color. Oecologia 1997, 109, 615-621. DOI: 10.1007/s004420050124. 52. Rahman, M.A.; Mak, R.; Ayad, H.; Smith, A.; Maclean, N. Expression of a novel piscine growth hormone gene results in growth enhancement in transgenic tilapia (Oreochromis niloticus). Transgenic Res. 1998, 7, 357-370. DOI: 10.1023/A:1008837105299 53. Ostenfeld, T.H.; McLean, E.; Devlin, R.H. Transgenesis change body and head shape in Pacific salmon. J. Fish. Biol. 1998, 52, 850-854. DOI: 10.1111/j.1095-8649.1998.tb00825.x. 54. Dunham, R.A.; Chatakondi, N.; Nichols, A.J.; Kucuktas, H.; Chen, T.T.; Powers, D.A.; Weetle, J.D.; Cummins, K.; Lovell, R.T. Effect of rainbow trout growth hormone complementary DNA on body shape, carcass yeld, and carcass composition of F1 and F2 transgenic common carp (Cyprinus carpio). Mar. Biotechnol. 2002, 4, 604-611. DOI: 10.1007/s10126-002-0034- 9. 55. Fares, N.H.; El-Sayed, A.K. Fine structural changes in the ileum of mice fed on endotoxin-treated potatoes and transgenic potatoes. Nat. Toxins. 1998, 6, 219-233. 56. Armon, S.; Yanai, O.; Sharon, E. Quantitative phenotyping of leaf margins in three dimensions, demonstrated on KNOTTED and TCP trangenics in Arabidopsis. J. Exp. Bot. 2014, 65, 2071- 2077, DOI: 10.1093/jxb/eru062. 57. Habes, D.; Messiad, R.; Gouasmia, S.; Grib, L. Effects of an inorganic insecticide (boric acid) against Blatella germanica: morphometric measurements and biochemical composition of ovaries. Afr. J. Biotechnol. 2013, 12, 2492-2497. 90

Appendix A

Figure A1. Experimental design scheme used for the captured individuals of C. quinquemaculatus in maize crops and adjacent forests of Atlantic Forest in the region of São Miguel do Oeste, SC, Brazil.

ARTIGO III: LOCAL EXTINCTIONS MAY BE EVIDENCED BY THE HOLES OF THE MORPHOMETRIC HYPERVOLUME IN DUNG BEETLES COMMUNITIES

Running head: Morphometric hypervolume in communities

Victor Michelon Alves1*, Malva Isabel Medina Hernández1

1 Programa de Pós Graduação em Ecologia, Departamento de Ecologia e Zoologia, Universidade Federal de Santa Catarina, Florianópolis, Santa Catarina, 88040–900

*Corresponding author Email: [email protected] (VMA)

Artigo submetido ao periódico Ecology

Local extinctions may be evidenced by the holes of the morphometric hypervolume in dung beetles communities

Abstract The body shape of a species is associated with its evolutionary history, reflecting behavioral peculiarities related to the ecological niche of each species. Thus, morphology can characterize the morphometric niche, and the species can be represented as body shape points within a morphometric universe. Taking into consideration the ecological community composed of these body shape points it is possible to calculate the morphometric diversity of the communities through hypervolume metrics, and it is also possible to calculate hole sizes that remain in this morphometric hypervolume, which are empty areas with no species. Such holes, representing non-filled spaces, may be “natural” or caused by a local extinction. In this study, we propose to evaluate the ecological community through the lenses of morphometric diversity. Therefore, we evaluated 38 dung beetle species from 30 subtropical communities in southern Brazil sampled in the summer of 2015, and included 15 forest remnant communities from the Atlantic Forest and 15 communities from adjacent maize cultivations. The shape of 495 specimens was measured using geometric morphometric techniques and using hypervolume we calculated the morphometric diversity and the hole size of each of the 30 communities. We found that the taxonomic diversity positively correlates with the morphometric diversity, and negatively correlates with the size of the holes. We verified that forest communities have higher values of morphometric diversity and smaller holes in the hypervolume than the maize cultivation communities, demonstrating that local extinction reduces community body shape spaces.

Keywords: Dung Beetles, Ecology, Geometric Morphometric, Morphology, Morphospace, Niche

Introduction The morphology of an organism is associated with its fitness and is shaped by its evolutionary history (Bergmann and Irschick 2010, Vidal-Garcia et al. 2014, Shao et al. 2016). Each species has a unique morphological design that is related to important aspects of its ecology, such as feeding (Michaux et al. 2007), habitat selection (Gaston et al. 2012), dispersion capacity (Johnson et al. 2015), predation (Burns et al. 2009), and thermal capacity (Miller and Denny 2011). In addition to being related to species ecology, morphological variation can also be a key issue in the structuring of natural communities (MacArthur and Levins 1967, Brown and Lieberman 1973, Stevens and Willig 2000). Due mainly to competition, dung beetle local communities are composed of a smaller portion of the regional pool of potentially colonizing species, with a limit on the number of species that coexist in a given community (Gardner et al., 2008, da Silva and Hernández 2014). Thus, coexistence between species is related to different methods of resource sharing (Palmer et al. 2003), and different strategies of resource use can be externally demonstrated by differences in body shape (Hernández et al. 2011). Furthermore, it is more likely that morphologically distinct species coexist with different forms of resource exploitation in a community (Brown and Bowers 1985), and it is expected that communities with more species may be more morphologically diverse. If the overlapping morphological design is too high, it is expected that such species, in depletion of resources, compete more directly within the community, or may differ in another dimension of the niche (e.g. nocturnal and diurnal habits) (see Soberón 2007). Some variables of the Eltonian niche, such as the use of resources, may be associated to an organism‟s morphology (Hernández et al. 2011); thus, the morphological design may be a reflection of the species‟ niche. Therefore, we assume that each species has its morphometric niche (MN), and we consider MN as the space occupied by a species within a particular morphometric universe. The morphometric diversity (MD) of a community can be calculated from the MN of each species, considering MD as the volume occupied by a community in the morphometric space, which can be measured through measuring hypervolume. The use of hypervolume metrics is very useful in detecting the total volume occupied by a community, so that communities with larger volumes would theoretically be more morphologically diverse. Another advantage is the detection of holes in the morphometric hypervolume, 95

such holes are the deviations of expectations, that is, empty areas in the hypervolume and that for probabilistic reasons could be occupied, and may reflect on ecological or evolutionary processes that are occurring outside the expected (Blonder 2016). Dung beetles are an ideal group for morphometric studies since they have different guilds that results in clear morphometric variations between the species (Hernández et al. 2011). These guilds are characterized by the habit associated with nesting, so the species use the same resource with different strategies, reducing direct competition and resulting in behavioral, ecological and morphometric variations (Halffter and Matthews 1966, Hernández et al. 2011). The tunneler dung beetle species, which feed on and build their nests through tunnels just below the food source, have an oval body with very little flattening, elongated elytra and contracted abdomen. The rollers, which roll food into round balls, are less oval and have slightly wider bodies. Dwellers, which eat and reside in the food resource, are elongated and flattened, and have a slight retracted cephalic portion (Hernández et al. 2011). In addition to the morphometric diversity, the dung beetle communities present high diversity of species, which respond quickly to environmental disturbances and are therefore used as ecological indicators (Halffter and Favila 1993). In this study, we propose to visualize the communities through the lens of morphometric diversity, starting with the description of the community through the occupation of each species in the body shape spaces within a morphometric universe. Thus, the objective of this study was to observe the relationship between taxonomic diversity and morphometric diversity in communities of dung beetles, and to compare the morphometric diversity and the hole size in the morphometric hypervolume of the communities coming from agricultural crops and forest remnants of the Atlantic Forest. Our hypotheses are: 1) there is a positive relationship between taxonomic diversity and morphometric diversity, since the decrease of species would reflect in the body shape spaces occupied by the species (morphometric niche); 2) communities in degraded areas (maize cultivation) have lower morphometric diversity and larger holes in the morphometric hypervolume, and such holes indicates body shape spaces that could be lost.

Materials and methods

Morphometric description of species We sampled dung beetles in 15 Atlantic Forest remnants and 15 adjacent maize cultivation areas in the region of São Miguel do Oeste (26°43'31''S, 53°31'05''W), state of Santa Catarina, southern Brazil, from December 2014 to February 2015. We worked with 38 species, of which 27 were captured in forest areas and 24 in cultivated areas. We measured 30 individuals per species (or the maximum number of individuals possible), totaling 495 individuals, currently deposited in the collection of the Federal University of Santa Catarina and the collection of the Federal University of Mato Grosso. In order to determine the body shape of each species, 15 landmarks were selected according to Hernández et al. (2011), where each landmark corresponds to a point in space defined by the three- dimensional Cartesian coordinates (X,Y,Z) that describe its geometric position in space (Bookstein 1991). Thus, the individuals were photographed in the dorsal, ventral and lateral sides, and the body shape was therefore captured in three dimensions (Fig. 1A-C). The 15 dung beetle landmarks were the following: 1) anterior margin of the clypeus; 2) ocular position; 3) division between pronotum and elytra; 4) division between thorax and abdomen; 5) posterior margin of the abdomen; 9) insertion point of the anterior legs; 10 and 11) insertion points of the central legs; 12) insertion point of the posteriors legs; 13) anterior point of convergence between the elytra; 14) central point of the elytra, and 15) posterior margin of the elytra. Points 6, 7, and 8 correspond to points 4, 3, and 2, respectively, for the other side of the body (Fig. 1D). The landmarks were captured using the program TpsDig 2.16 (Rohlf 2010) and tpsUtil 1.53 (Rohlf 2012).

Fig. 1. Landmarks (red points) used in the construction of the body shape. (A) Dorsal view; (B) Ventral view; (C) Lateral view; (D) Graphical representation of body shape based on the 15 landmarks in three dimensions, adapted from Hernández et al. 2011.

For the shape analysis of each species, the coordinates of the landmarks were submitted to a Generalized Procrustes analysis (GPA). This analysis is a multivariate technique that detects variations between configurations (in this case the different shapes of the insect‟s body), involving transformations in three steps: (1) first the effect of the position is removed by superimposing the configuration of the shapes; (2) the effect of size is eliminated, proportioning the size of the centroid; (3) and lastly the configurations are rotated, thus eliminating the orientation effect (Rohlf and Slice 1990). After these steps, a mean configuration of the species structure (consensus configuration) is provided and the variation in shape can be evaluated among the species studied. In order to visualize the position of the species within the multivariate space, a Principal Component Analysis (PCA) was performed. The body shape analyzes for each species (GPA and PCA) were conducted in the software Morpho J 1.05d (Klingenberg 2011).

Construction of the communities hypervolumes In order to construct the hypervolumes representing the community‟s body shape spaces or morphometric diversity, the average position of each species was used in the first three axes of the PCA, thus a global hypervolume was constructed in three dimensions, including the values of the 38 species (regional pool or “global community”). Subsequently, a hypervolume was created for each of the 30 local communities, in order to compare the volume of the hypervolume between each other. We consider the volume of the hypervolume of each community as the morphometric diversity value. An important methodological step in this comparison is that the hypervolumes have the same bandwidth, that is, they can be “sliced” at the same height within the volume. Therefore, the bandwidth was estimated using the estimator plug-in with a quantile threshold of 0% (Blonder et al. 2014) of the global community (with 38 species). The global community bandwidth was used for each local community, so all communities had the same bandwidth value and their volumes could be compared. In order to measure the volume of the holes in the hypervolume structure of each of the 30 local communities, the difference between the expectation box (see Blonder et al. 2014) of the global community and the observed hypervolume of each local community was used. Thus, it was possible to identify which regions of the morphometric hypervolume are lost in each local community. As the hypervolume follows a multidimensional kernel density estimation procedure, we calculate each hypervolume volume and each hole ten times. The coefficient of variation (CV) were low for both hypervolume volume (0.3 – 0.4%) and holes sizes (1-7%) (Appendix S1, Table S1). Thus, average values of the hypervolume volume and holes area calculated taking into account these ten repetitions (Table S1). For hypervolume analysis, the hypervolume package was used (Blonder et al. 2014) in the R software (R Core Team, 2015).

Statistical Analysis of the data The relationship between taxonomic diversity (species richness) and morphometric diversity (volume of hypervolume) was verified through a regression analysis and an analysis of covariance (ANCOVA) was performed to test if morphometric diversity was affected by the area (crops or forest fragments). For that, the response variable was logarithmized (ln). Using the same analysis, the relationship between taxonomic diversity (species richness) and the size of the holes in the hypervolume was tested, including both areas (crops or forest 99

fragments). Regression analysis and ANCOVA were performed in R software (R Core Team, 2015).

Results Species morphometrics and construction of the morphometric universe The morphometric universe of the global community, which includes 38 points in the body shape spaces representing the morphometric niche of each species (see the 38 points of "heads of pins" in Fig 2A), was associated with body shape variations of species depending on behavior and evolutionary history. Many species present high positive values in PC1, reflecting morphometric variations associated with the habit of constructing tunnels, which have a more robust and oval shaped body, with a slight retraction in the anteroposterior axis and enlargement in the dorsoventral axis, mainly between the elytra and the abdomen. Typical genera that have this body shape are Canthidium and Dichotomius, represented to the right of Fig. 2A. The species observed on the opposite side are isolated in the morphometric universe with negative values in PC1 and PC3, and are represented by the genus Eurysternus, that unlike tunneler species, have strong dorsoventral flattening and are elongated in the anteroposterior axis, such morphometric variations are probably associated with the nesting habit of dwellers. The species with rolling behavior have greater body shape plasticity, occupying more spaces within the morphometric universe. This is the case of the representatives of the genus Canthon, wich have positive values in PC1 (a reflection of the slightly oval body shape and enlarged abdominal region) and Deltochilum, with positive positions in PC2 and PC3, a result of the oval body with a strong retraction of the cephalic portion. Other genera found in this study occupy intermediate positions in the morphometric universe (Appendix S1, Table S2).

100

Fig. 2. Morphometric universe and global hypervolume. (A) Global morphometric universe of dung beetles, with each of the 38 species occupying a specific region within the universe, some genera were drawn to represent the variation in body shape; (B) hypervolume of the global community (red points) composed by 38 species and hypervolume holes (blue dots).

The morphometric hypervolume was formed by 38 species from the “global community” (red points Fig 2B), leaving holes in the middle of this hypervolume (blue dots Fig 2B), these holes are probably naturally empty areas, that is, from non-existent body shapes. There are several species that have close morphometric niches, however, it is important to take into account that very close morphological designs do not necessarily reflect in an ecological niche overlap. The same designs can inhabit very different conditions, where species are distanced by other variables of Eltonian or Grinnellian niche. Furthermore, each point in Fig. 2A corresponds to an average of each species; however, observing the position occupied by all individuals from each species, an intraspecific body shape variation is observed (Fig 3). Thus, this variation shows what space occupied by all individuals of a species overlaps with the space occupied by other species, therefore creating a morphometric overlap. Such is the case for the individuals measured from Dichotomius sericeus, which overlap with the individuals from D. nisus (represented in Fig. 3 by the two first axes in the PCA), this overlap would lead us to think that these species occupy the same niche in the local community, however, when analyzing spatial distribution we found that the species D. sericeus is mostly found in forest areas while D. nisus is found in open areas. Similarly, Deltochilum brasiliense and D. morbillosum are forest species extremely similar in body shape, but differ in relation to body size; the body size is an important variable in niche differentiation in dung 101

beetles (Hernández et al. 2011). Another example of morphometric proximity are the individuals Eurysternus parallelus and E. caribaeus that present a substantial morphological overlap, where both are from forest areas, however, there are temporal differences in the use of resources, since E. caribaeus is cited as nocturnal (Barragán et al. 2011) while E. parallelus is preferentially diurnal (Hernández 2002) (Fig. 3).

Fig. 3. Morphometric niche represented in the first two axes of the Principal Component Analysis. Drawn are examples of morphometric niche and width of morphometric niche (95% C.I.) of six species that have overlapping of their morphometric niches, however, are differentiated in other ecological niche variables.

Morphometric diversity and hypervolume hole size The 15 dung beetles communities from the forest fragments have higher morphometric diversity values than the communities from cultivation areas, since the volume of hypervolumes were larger, varying between 0.0025 and 0.0058 (Table S1, red points Fig 4), while the cultivation communities varied between 0.0022 and 0.0050 (Table S1, red points Fig 5). In contrast, the volume of the holes was smaller in the forest fragment communities (Table S1, green points Fig 4), varying between 102

0.0001 and 0.0009, than in the cultivation areas that ranged between 0.0002 and 0.00156 (Table S1, green points Fig 5). Unlike the holes in the “global community”, these holes are not natural and reflect the loss of body shape spaces as a consequence of agricultural conversion.

Fig. 4. Morphometric diversity (red points) and holes (green points) of the 15 dung beetle communities from the forest fragments, collected in the region São Miguel do Oeste, SC, Brazil. The hypervolume volume was considered as the value of morphometric diversity

Fig. 5. Morphometric diversity (red points) and holes (green points) of the 15 dung beetle communities from maize cultivation areas, collected in the region São Miguel do Oeste, SC, Brazil. The hypervolume volume was considered as the value of morphometric diversity.

103

Relation between taxonomic diversity, morphometric diversity and hole size There is a positive relationship between species diversity and morphometric diversity for both, forest and agricultural crops communities (ANCOVA p <0.01, Appendix S1, Table S3), showing that richer communities are more morphometrically diversified (Fig. 6A). This is plausible since, beginning with the principle of the morphometric niche, each species occupies a different position in the morphometric universe, and the sum of these occupations will result in greater morphometric diversity. Since the relation is curvilinear, the morphometric diversity has a lower rate of growth in a richer community (Fig. 6A). When new species are added to the community (increase in X values), the morphometric diversity will increase at a lower rate due to the fact that communities would tend to a “morphometric saturation”, and the addition of new species would only result in overlapping morphometric niches. The volume of the morphometric hypevolume holes is also related to species richness, and their relation is negative: more diverse communities have smaller holes (Fig. 6B). Both, forest and agricultural crops communities, showing that richer communities have smaller holes in the hypervolume (ANCOVA p <0.01, Appendix S1,Table S4). The relationship is also curvilinear due to the fact that as the community becomes richer, the decrease rate of holes size is lower (Fig 6 B). In other words, in rich communities, even with the addition of more species in the community, the hole is not completely filled, due to the existence of the “natural hole”. However, in cultivation communities, besides the existence of the “natural hole” there are holes caused by anthropic action, with the loss of body shape spaces (species) that were locally extinct in these areas.

104

Fig. 6. Relationship among species diversity, morphometric diversity and hypervolume hole size. (A) Relationship between morphometric diversity to species diversity; (B) relationship of hole size to species diversity in 30 communities of dung beetles, 15 forest communities and 15 maize cultivation communities, collected in São Miguel do Oeste, SC, Brazil.

105

Discussion In this study we described the communities as a total shape space which is constructed when species structuring the communities. This new communities representation was made with the union of two important tools: geometric morphometrics and hypervolume. Using this approach we were able to calculate important information, such as the morphometric niche and morphometric diversity. Thus, we were able to describe the community through a different point of view, with each species representing a point in space in the body shape universe. We were also able to demonstrate that there are “natural holes” and holes caused by anthropic action, and that richer communities are more morphometrically diversified and have smaller holes in the morphometric hypervolume. Althoug in dung beetles the application of morphometry has been used to elucidate phylogenetic and evolutionary relationships (Pretorius et al. 2000, Bai et al. 2011, Bai et al. 2015), species differentiation (Pizzo et al. 2009), and ecomorphological studies (Inward et al. 2011), our work presents a completely different perspective when viewing the community from a totally morphometric point of view. Hernández and collaborators (2011) were the first to apply geometric morphometrics in describing dung beetle communities by describing a “global community” composed of 39 species. In our study we describe the morphometric universe of 38 species, and found a pattern similar to Hernández (2011), with the main variations in body shape related to the behavior of resource allocation. Species of the genus Eurysternus were isolated in morphospace, presenting characteristics that reflect their behavior as dwellers (species that do not bury the resource). However, since methods for accessing hypervolume have become accessible recently (Blonder et al. 2014), we have broadened the proposal of Hernández et al. (2011) calculating the morphometric diversity and hole size in the morphospace. The main result of the present study was to describe the positive relationship between taxonomic and morphometric diversity. This is probably a reflection of resource use specialization, which results in greater morphological variation (Maglianesi et al. 2014) and as the environment provides a greater variety of resources, the niche can be expanded. Within the most diverse communities, the increase in species richness would result in overlapping of the morphometric niche and the differentiation could be made using other dimensions of the ecological niche. In the communities analyzed in this study we could show that some species, very close morphologically, presented spatial differentiations (forest areas and open areas) and others presented 106

temporal differentiations (diurnal and nocturnal habit). The difference in activity time of the species is a classic example of resource sharing (Hernández 2002), as well as resource specialization (Bedoussac et al. 2007). In very diverse environments the specialization by resource is more evident, and allows an increase in the number of species without altering the morphometric diversity. For example, in tropical regions there are species of dung beetles that live on the coat of some species of mammals, such as sloths (Boucomont 1928, Ratcliffe 1980). It is also possible that species that present a similar shape are different in sizes, which increases the possibilities of obtaining resources and decreases the possible competition among species (Hernández et al. 2011). We also observed in this study that the higher the species richness of a community, the smaller the number of holes in the hypervolume, which is due to filling of body shape spaces by new species being added to the community. However, there is the “natural” limitation in filling the space, since there are body shapes that do not exist and thus the morphometric space will not be completely filled. The existence of “natural holes” in the hypervolume, which represents parts of space that were not occupied throughout evolutionary time, was raised by Blonder (2016) who developed mathematical tools for their calculations. In the case of the dung beetles the natural holes are detected in different positions, including in intermediate spaces between the genera Eurysternus and several genera of rollers. This hole may have been caused because the intermediate shape has not been positively selected (Arnold 1992) or it may be an area of “evolutionary void”, that is, there are still possibilities for this hole to be filled. The natural hole of the morphometric hypervolume can represent the vacant niche, where there are no adapted species for specific environmental conditions (Lawton 1984, Cornell 1999), although the existence of vacant niches is extremely questioned, since niche is a property of the organism and not the environment (Chase and Leibold 2003). Therefore, considering the niche as a property of the species, as several researchers argue, the morphometric hole could be filled by ecological and evolutionary processes. Since the body shape of dung beetles is strongly related to the behavioral guild (Hernández et al. 2011), the hole could be occupied by species that are not found in the study region and cannot be classified within the three behavioral guilds traditionally described (rollers, tunnelers and dwellers). For example, there are dung beetles from the Eucraniini tribe that collect the resource with their front legs, and the posterior and middle legs are for moving to feed or nest (Ocampo and Philips 2005). This type of behavior may 107

require a different phenotype that could fill the natural hole. Moreover, a small behavioral variation could alter the individual‟s shape (due to differences in insertions or muscle development) and fill the natural hole. For example, specimens of the genus Pachysoma use their posterior legs to collect resource, and with their anterior legs they move forward and not backwards, as is done in common rollers (Ocampo and Philips 2005). We found holes caused by anthropic action, which represent body shape spaces that were locally extinct when the forest areas were converted into cultivation areas. The loss of body shape spaces, in agricultural areas, was mainly due to Deltochilum genus local extinction. These species are large rollers (more than 1 cm) with a peculiar morphology: strong cephalic retraction with pronotum and elytra curved, resulting in a convex body shape. Physiological studies have shown that Deltochilum brasiliense have a fast rate of heat acquisition (Amore et al. 2017), therefore in areas with high solar radiation intensity (such as agricultural crops) individuals could overheat. We know that in dung beetles the largest species are lost when natural areas are converted into monocultures (Gardner et al. 2008, Horgan 2008), even we found large species in the corn crops (e.g. Dichotomius nisus, larger tunneler). Therefore, the relationship between body size and characteristics associated with the shape emerges as a good predictor of species extinction in open areas. We observed that crop communities holes are located in similar places to each other (Fig. 5). This pattern may be caused because crops communities have a smaller shape space to be occupied (e.g. because Deltochilum local extinction) reflecting the smaller possibilities of use of the morphological space when deforestation occurs. In contrast, the forest communities holes are located in different places of the morphospace (Fig. 4). The forest communities have a greater shape space that could be occupied, and the different structures of the communities reflect the several possibilities of occupying this shape space. Also, community morphospace reduction shows the ecosystem functions loss in crop areas, since from a functional point of view (Tilman 2001) the morphometric niche may be visualizes as the functional niche (Rosenfeld 2002). Therefore, the larger morphometric holes in crop communities represent both the reduction of taxonomic and the reduction of functional diversity in degraded areas.

108

Conclusions In conclusion, our study provides a morphological description of natural communities. With this new approach we can describe a community composed of points in the space of a shape universe. The positive relationship between species richness and morphometric diversity may be associated with the niche partition within the community. The holes found in the morphometric hypervolume prove the loss of species in the anthropic areas (corn crops), whereas the natural holes represent phenotypes that probably were not positively selected. Our results demonstrate the applicability of our approach that can be used in the study of other natural communities.

Acknowledgments We are grateful to Dr. Fernando Vaz-de-Mello (Federal University of Mato Grosso, Brazil) for dung beetle identification and Dr. Benjamin Blonder for his help regarding the use of the hypervolume package. We are grateful to Dr. Paulo Emilio Lovato (Department of Rural Engineering /UFSC) for coordinating the project “Fortalecimento das condições de produção e oferta de sementes de milho para a produção orgânica e agroecológica do Sul do Brasil” (CNPq Call 048/13). VMA thanks CAPES (Ministry of Education of Brazil) for a PhD Grant and MIMH thanks CNPq (Proc. 309030/2013-7) for a Research Productivity Grant.

References Amore, V., M.I.M. Hernandez, L.M. Carrascal, and J.M. Lobo. 2017. Exoskeleton may influence the internal body temperatures of Neotropical dung beetles (Col. Scarabaeinae). PeerJ 5: e3349. Arnold, S.J. 1992. Constraints on phenotypic evolution. The American Naturalist 140:85-107. Bai, M., E. McCullough, K.Q. Song, W.G. Liu, and X.K. Yang. 2011. Evolutionary constraints in hind wing shape in chinese dung beetles (Coleoptera: Scarabaeinae). PLoS ONE 6:e21600. Bai, M., S. Li, Y. Lu, H. Yang, Y. Tong, and X. Yang. 2015. Mandible evolution in the Scarabaeinae (Coleoptera: Scarabaeidae) and adaptations to coprophagous habits. Frontiers in Zoology 12:30. 109

Barragán, F., C.E. Moreno, F. Escobar, G. Halffter, and D. Navarrete. 2011. Negative impacts of human land use on dung beetle functional diversity. PLoS ONE 6:e17976. Bedoussac, L., M.E. Favila, and R.M. López. 2007. Defensive volatile secretions of two diplopod species attract the carrion ball roller scarab Canthon morsei (Coleoptera: Scarabaeidae). Chemoecology 17:163–167. Bergmann, P.J., and D.J. Irschick. 2010. Alternate pathways of body shape evolution translate into common patterns of locomotor evolution in two clades of lizards. Evolution 64:1569–1582. Blonder, B., C. Lamanna, C. Violle, and B.J. Enquist. 2014. The n- dimensional hypervolume. Global Ecology and Biogeography 23:595–609. Blonder, B. 2016. Do hypervolumes have holes? The American Naturalist 187:93-105. Bookstein, F.L. 1991. Morphometric tools for landmark data: Geometry and Biology. Cambridge University Press, Cambridge, UK. Boucomont, A. 1928. Coprophages d'Amerique du Sud nouveaux ou peu connus. Bulletin de la Société entomologique de France 12:186-200. Brown, J.H., and G.A. Lieberman. 1973. Resource utilization and coexistence of seed-eating desert rodents in sand dune habitats. Ecology 54: 788-797. Brown, J.H., and M.A. Bowers. 1985. Community organization in hummingbirds: relationships between morphology and ecology. The Auk 102: 251-269. Burns, J.G., P. Di Nardo and F.H. Rodd. 2009. The role of predation in variation in body shape in guppies Poecilia reticulata: a comparison of field and common garden phenotypes. Journal of Fish Biology 75: 1144-1157. Chase, J.M., and M.A. Leibold. 2003. Ecological Niches: Linking Classical and Contemporary Approaches. University of Chicago Press, Chicago, USA. Cornell, H.V. 1999. Unsaturation and regional influences on species richness in ecological communities: a review of the evidence. Écoscience 6: 303–315. da Silva, P.G., and M.I.M. Hernández. 2014. Local and regional effects on community structure of dung beetles in a mainland- island scenario. PLoS ONE 9: e111883. 110

Gardner, T.A., M.I.M. Hernández, J. Barlow, and C.A. Peres. 2008. Understanding the biodiversity consequences of habitat change: the value of secondary and plantation forests for neotropical dung beetles. Journal of Applied Ecology 45: 883- 893. Gaston, K.A., J.A. Eft, and T.E. Lauer. 2012. Morphology and its effect on habitat selection of stream fishes. Proceedings of the Indiana Academy of Science 121: 71-78. Halffter, G., and E.G. Matthews. 1966. The natural history of dung beetles of the subfamily Scarabaeinae (Coleoptera: Scarabaeidae). Folia Entomológica Mexicana 12: 1-312. Halffter, G., and M.E. Favila. 1993. The Scarabaeinae (Insecta: Coleoptera) an animal group for analysing, inventorying and monitoring biodiversity in Tropical Rainforest and modified landscapes. Biology International 27: 15–21. Hernández, M.I.M. 2002. The night and day of dung beetles (Coleoptera, Scarabaeidae) in the Serra do Japi, Brazil: elytra colour related to daily activity. Revista Brasileira de Entomologia 46: 597-600. Hernández, M.I.M., L.R. Monteiro, and M.E. Favila. 2011. The role of body size and shape in understanding competitive interactions within a community of Neotropical dung beetles. Journal of Insect Science 11: 1-14. Horgan, F.G. 2008. Dung beetle assemblages in forests and pastures of El Salvador: a functional comparison. Biodiversity and Conservation 17: 2961–2978. Inward, D.J.G., R.G. Davies, C. Pergande, A.J. Denham, and A.P. Vogler. 2011. Local and regional ecological morphology of dung beetle assemblages across four biogeographic regions. Journal of Biogeography 38: 1668–1682. Johnson, J.B., D. Saenz, C.K. Adams, and T.J. Hibbitts. 2015. Naturally occurring variation in tadpole morphology and performance linked to predator regime. Ecology and Evolution 5: 2991-3002. Klingenberg, C.P. 2011. MorphoJ: an integrated software package for geometric morphometrics. Molecular Ecology Resources 11: 353-357. Lawton, J.H. 1984. Non-competitive populations, non-convergent communities, and vacant niches: the herbivores of bracken. In: Strong, D.R., D. Simberloff, L.G. Abele and A.B. Thistle 111

(Eds.), Ecological Communities: Conceptual Issues and the Evidence. Princeton University Press, pp. 67-101. MacArthur, R., and R. Levins. 1967. The limiting similarity, convergence and divergence in coexisting species. The American Naturalist 101: 377–385. Maglianesi, M.A., N. Blüthgen, K. Böhning-Gaese, and M. Schleuning. 2014. Morphological traits determine specialization and resource use in plant–hummingbird networks in the neotropics. Ecology 95: 3325-3334. Michaux, J., P. Chevret, and S. Renaud. 2007. Morphological diversity of Old World rats and mice (Rodentia, Muridae) mandible in relation with phylogeny and adaptation. Journal of Zoological Systematics and Evolutionary Research 45: 263- 279. Miller, L.P., and M.W. Denny. 2011. Importance of behavior and morphological traits for controlling body temperature in littorinid snails. The Biological Bulletin 220: 209-223. Ocampo, F.C., and T.K. Philips. 2005. Food relocation behavior of the Argentinian dung beetle genus Eucranium Brullé and comparison with the southwest African Scarabaeus (Pachysoma) MacLeay (Coleoptera: Scarabaeidae: Scarabaeinae). Revista de la Sociedad Entomológica Argentina 64: 53-59. Palmer, T.M., M.L. Stanton, and T.P. Young. 2003. Competition and coexistence: exploring mechanisms that restrict and maintain diversity within mutualist guilds. The American Naturalist 162: 63–79. Pizzo, A., A.L.M. Macagno, A. Roggero, A. Rolando, and C. Palestrini. 2009. Epipharynx shape as a tool to reveal differentiation patterns between insect sister species insights: from Onthophagus taurus and O. illyricus (Insecta: Coleoptera: Scarabaeidae). Organisms Diversity & Evolution 9: 189-200. Pretorius, R., K. Philips, and C. Scholtz. 2000. Geometric morphometrics, the metendosternite and its use in phylogenetics of the Scarabaeinae (Coleoptera). Elytron 14: 125-148. R Core Team. 2015. R: A language and environment for statistical computing. R Foundation for Statistical Computing, . 112

Ratcliffe, B.C. 1980. New species of coprini (Coleoptera: Scarabaeidae: Scarabaeinae) taken from the pelage of three toed sloths (Bradypus tridactylus l.) (Edentata: Bradypodidae) in central amazonia with a brief commentary on scarab-sloth relationships. The Coleopterists Bulletin 34: 337-350. Rohlf, F.J., and D. Slice. 1990. Extensions of the Procrustes method for the optimal superimposition of landmarks. Systematic Zoology 39: 40-59. Rohlf, F.J. 2010. TpsDig, Version 2.16. Department of Ecology and Evolution, State University of New York at Stony Brook, . Rohlf, F.J. 2010. TpsUtil, Version 1.53. Department of Ecology and Evolution, State University of New York at Stony Brook, . Rosenfeld, J.S. 2002. Functional redudancy in ecology and conservation. Oikos 98: 156-162. Shao, S., Q. Quan, T. Cai, G. Song, Y. Qu, and F. Lei. 2016. Evolution of body morphology and beak shape revealed by a morphometric analysis of 14 Paridae species. Frontiers in Zoology 13: 30. Soberón, J. 2007. Grinnellian and Eltonian niches and geographic distributions of species. Ecology Letters 10: 1115–1123. Stevens, R.D., and M.R. Willig. 2000 Community structure, abundance and morphology. Oikos 88: 48-56. Tilman, D. 2001. Functional Diversity. In: S.A. Levin (Ed.) Encyclopedia of Biodiversity. Academic Press, New York. Vidal-Garcia, M., P.G. Byrne, J.D. Roberts, and J.S. Keogh. 2014. The role of phylogeny and ecology in shaping morphology in 21 genera and 127 species of Australo-Papuan myobatrachid frogs. Journal of Evolutionary Biology 27: 181-192.

113

Appendix S1

Table S1. Values of morphometric diversity (hypervolume volume, HV) and hole size (HS) followed by confidence intervals (CI) and coefficients of variation (CV). Forest fragment communities (FF); maize crops communities (MC). Area HV CI CV (%) HS CI CV (%) FF (1) 0.005449251 0.005460765 - 0.005437737 0.34 0.000216779 0.000225002 - 0.000208557 6.11 FF (2) 0.005144394 0.005153807 - 0.005134981 0.29 0.000270802 0.000283512 - 0.000258091 7.52 FF (3) 0.002521049 0.002526302 - 0.002515796 0.33 0.000912555 0.000925647 - 0.00899463 2.31 FF (4) 0.004118838 0.004125075 - 0.004112600 0.24 0.000368619 0.000378794 - 0.000358444 4.45 FF (5) 0.004817820 0.004825947 - 0.004809694 0.27 0.000298766 0.000313539 - 0.000283992 7.97 FF (6) 0.005845145 0.005861087 - 0.005829202 0.44 0.000171316 0.000182741 - 0.000159891 9.76 FF (7) 0.004849363 0.004858492 - 0.004840235 0.30 0.000314678 0.000330202 - 0.000299155 7.95 FF (8) 0.002713431 0.002720096 - 0.002706767 0.39 0.00068468 0.000694421 - 0.000674938 2.29 FF (9) 0.003773723 0.003780798 - 0.003766649 0.30 0.000430804 0.00043751 - 0.000424098 2.51 FF (10) 0.003892982 0.003900767 - 0.003885198 0.32 0.000459195 0.000467557 - 0.000450833 2.93 FF (11) 0.004521099 0.004528837 - 0.00451336 0.27 0.000409061 0.000416667 - 0.000401455 3.00 FF (12) 0.004635282 0.004645337 - 0.004625227 0.35 0.000331574 0.000343916 - 0.000319232 6.00 FF (13) 0.003386819 0.003390766 - 0.003382871 0.18 0.000536275 0.000545939 - 0.000526611 2.90 FF (14) 0.003921953 0.003929441 - 0.003914464 0.30 0.000489156 0.000498237 - 0.000480075 3.00 FF (15) 0.003635400 0.003642591 - 0.003628209 0.32 0.000491166 0.000499707 - 0.000482625 2.80 114

MC (1) 0.004001089 0.004011193 - 0.003990985 0.40 0.000649893 0.000663647 - 0.000636139 3.41 MC (2) 0.00500337 0.005013723 - 0.004993018 0.33 0.000279043 0.000292222 - 0.000265863 7.62 MC (3) 0.002271706 0.002276948 - 0.002266463 0.37 0.001561006 0.001576437 - 0.001545574 1.60 MC (4) 0.003057588 0.003062693 - 0.003052483 0.30 0.000933593 0.000950137 - 0.000917049 2.86 MC (5) 0.00313714 0.003141383 - 0.003132897 0.22 0.000995207 0.001005348 - 0.000985066 1.64 MC (6) 0.003430588 0.003437888 - 0.003423288 0.34 0.000661984 0.000677396 - 0.000646572 3.75 MC (7) 0.002396638 0.002401983 - 0.002391294 0.36 0.000676194 0.000683427 - 0.000668961 1.72 MC (8) 0.002593863 0.00260054 - 0.002587185 0.41 0.001022694 0.00104435 - 0.001001038 3.41 MC (9) 0.003332491 0.003336817 - 0.003328166 0.21 0.000678552 0.000690795 - 0.000666309 2.91 MC (10) 0.002635113 0.002640199 - 0.002630028 0.31 0.000876039 0.000888844 - 0.000863234 2.36 MC (11) 0.002805581 0.002810913 - 0.002800249 0.30 0.000798872 0.000808529 - 0.000789215 1.95 MC (12) 0.0027768 0.002780936 - 0.002772663 0.24 0.001096867 0.001101937 - 0.001091798 0.74 MC (13) 0.002430869 0.00243961 - 0.002422127 0.58 0.000810757 0.000816749 - 0.000804764 1.19 MC (14) 0.002786216 0.002791321 - 0.002781111 0.29 0.000580731 0.000586808 - 0.000574653 1.65 MC (15) 0.002355301 0.002362377 - 0.002348225 0.48 0.000922522 0.000930551 - 0.000914493 1.40 115

Table S2. Dung beetle species collected in the region of São Miguel do Oeste, SC, Brazil. Body shape values in the first three principal component axes (PC1, PC2, and PC3).

Species PC1 PC2 PC3 Anomiopus sp.1 -0.0661 -0.0430 0.0155 Canthidium cavifrons 0.0981 0.0080 0.0494 Canthidium dispar 0.0526 -0.0138 -0.0007 Canthidium moestum 0.0308 -0.0432 0.0029 Canthidium aff. trinodosum 0.0575 -0.0159 -0.0082 Canthidium sp. 1 0.0479 -0.0616 0.0088 Canthidium sp. 2 0.0366 -0.1025 -0.0456 Canthidium sp. 3 0.0804 -0.0369 -0.0224 Canthon angularis 0.0441 -0.0069 0.0101 Canthon chalybaeus 0.0060 0.0072 0.0218 Canthon laminatus 0.0170 -0.0216 0.0942 Canthon lividus lividus 0.0019 0.0043 0.0210 Canthon podagricus -0.0072 -0.0605 0.0635 Canthon quinquemaculatus 0.0464 -0.0191 0.0083 Canthon tetraodon 0.0290 -0.0416 0.0112 Canthon aff. lituratus 0.0202 -0.0832 -0.0103 Canthon aff. luctuosus 0.0000 0.0110 0.0296 Canthon aff. mutabilis 0.0172 -0.0795 0.0179 Canthonella catharinensis -0.0190 0.0186 -0.0859 Canthonella aff. instriata 0.0096 0.1077 -0.0143 Coprophanaeus saphirinus -0.0307 -0.0397 -0.0192 Deltochilum brasiliense -0.0272 0.0832 0.0097 Deltochilum morbillosum -0.0327 0.0834 0.0164 Deltochilum rubripenne -0.0423 0.0475 0.0501 Dichotomius bicuspis -0.0047 -0.0054 0.0682 Dichotomius nisus 0.0565 0.0376 -0.0529 Dichotomius sericeus 0.0512 0.0304 -0.0219 Dichotomius sp.1 0.0635 -0.0167 -0.0030 Eurysternus aeneus -0.1125 0.0042 -0.0239 Eurysternus caribaeus -0.1625 -0.0005 -0.0184 Eurysternus parallelus -0.1450 -0.0092 -0.0238 Ontherus sulcator 0.0374 -0.0203 -0.0052 Onthophagus aff. hirculus 0.0215 -0.1167 0.0035 Pseudocanthon sp.1 0.0019 0.0041 0.0237 Scybalocanthon nigriceps 0.0267 0.0509 -0.0170 Sulcophanaeus menelas -0.0087 -0.0222 -0.0170 Trichillum externepunctatum 0.0560 0.0465 0.0373 Uroxy dilaticollis 0.0191 0.0145 0.0455 116

Table S3. ANCOVA results of the variables that affect the morphometric diversity (in ln) of 30 dung beetles communities (15 forest communities and 15 maize cultivation communities) collected in São Miguel do Oeste, SC, Brazil. DF Sum of Mean F p squares squares Species diversity 1 1.76 1.76 108.951 <0.001 Area (maize crop or forest) 1 0.10 0.10 6.15 0.02 Species diversity * Area 1 0.03 0.03 1.77 0.19 Residuals 26 0.42 0.02

117

Table S4. ANCOVA results of the variables that affect the morphometric holes size (in ln) of 30 dung beetles communities, 15 forest communities and 15 maize cultivation communities, collected in São Miguel do Oeste, SC, Brazil. DF Sum of Mean F p squares squares Species diversity 1 6.74 6.74 143.839 <0.001 Area (maize crop or forest) 1 0.43 0.43 9.16 <0.01 Species diversity * Area 1 0.05 0.05 1.05 0.31 Residuals 26 1.22 0.05

118

119

ARTIGO IV: ELYTRA ABSORB ULTRAVIOLET RADIATION BUT TRANSMIT INFRARED RADIATION IN NEOTROPICAL CANTHON SPECIES (COLEOPTERA, SCARABAEINAE)

Victor Michelon Alves1, Malva I. M. Hernandez1, Jorge M. Lobo*2

1 Programa de Pós Graduação em Ecologia, Departamento de Ecologia e Zoologia, Universidade Federal de Santa Catarina, Florianópolis, Santa Catarina, 88040–900

2 Department of Biogeography and Global Change, Museo Nacional de Ciencias Naturales, CSIC, Madrid, Spain

*Correspondence author: [email protected] (Jorge M. Lobo)

Artigo aceito para publicação no periódico Photochemistry and Photobiology

120

121

Elytra absorb ultraviolet radiation but transmit infrared radiation in Neotropical Canthon species (Coleoptera, Scarabaeinae)

Abstract Strategies to deal with global radiation may be related to important aspects of species biology and ecology by reflecting, transmitting or absorbing the radiation of varying wavelengths differently. The elytra capacity to manage infrared, visible and ultraviolet radiations (from 185 to 1400 nm) was assessed with a spectrophotometric analysis in five Canthon species of dung beetles; we calculated the reflectance, transmittance, and absorbance capacity of the elytra of these species. These species have different ecologies, since: two species preferentially inhabit forest areas (Canthon angularis and Canthon lividus lividus), two species preferentially inhabit open areas (Canthon chalybaeus and Canthon tetraodon) including agricultural crops, and one species does not present a clear habitat preference and can be found in both habitats (Canthon quinquemaculatus). All the species show a similar pattern in which the light from shorter wavelengths and higher frequencies is almost entirely absorbed by the elytra, while radiation from longer wavelengths and lower frequencies can mostly pass through the elytra. However, C. quinquemaculatus seems to have significantly higher rates of reflectance and transmittance in the visible and near-infrared spectrum. This different pattern found in C. quinquemaculatus may be associated with its capacity to establish populations both in agricultural and forest areas.

122

Introduction The sun is the major source of radiation on Earth and plays a crucial role in the evolution of life (1). The sunlight reaching Earth‟s surface is called global radiation (2), and is frequently classified into three ranges (based on human vision) for biological studies according to their electromagnetic wavelength: infrared, visible and ultraviolet radiation. These radiations may affect all living organisms. Distinct radiations have different energies, and energy increases as the wavelength decreases. Thus, species capacity to deal with each type of radiation - reflecting, absorbing or transmitting it - may affect important aspects of their biology. In insects, transmitting infrared radiation into the body can increase internal body temperature, influencing their development (3), immune system regulation (4), fertility (5), foraging habit (6) and morphology (7). Reflecting infrared radiation can prevent overheating, which helps individuals to maintain optimal temperatures, as high temperatures (mostly over 50ºC) are fatal for many insects, as well as for dung beetles (8). Visible radiation is responsible for the abundant color variation observed in insects and is an important feature interfering in species behavior, ecology and physiology. Coloration patterns may be associated with the quality of the immune system (9), which can influence mate choice and consequently sexual selection (10). Other evolutionary processes may drive coloration patterns; especially important, is the pressure of predation, where individuals develop color patterns for camouflage (11), as a warning (aposematic) (12) or mimicry. In Coleoptera (13), and specifically in dung beetles (14, 15), color variations cannot be pigmentary, but structural. Thus, beetle's color is understood as a pure physical mechanism due to material properties of the exoskeleton which reflects, scatters and deflects the light. Coloration also has a possible role in thermoregulation (16). Experiments with the grasshopper Kosciuscola tristis showed that increasing the ambient temperature (from 15ºC to 25ºC) activates intracellular granule migration that changes insect color from black to blue (17-18). This color change has a thermoregulatory function, since the black coloration is associated with heat absorption at lower temperatures (19). Therefore, coloration presents two distinct aspects: ecological (associated with species interactions) and physiological (related to species homeostasis). These two evolutionary paths suggest the existence of a trade-off between ecological and physiological investment. Furthermore, coloration can be associated with species habits, for example, in tropical dung beetles black species are usually 123

nocturnal while diurnal species are colorful (20). This pattern was only observed in large species (> 1 cm) as small species may be black and diurnal (20). Thus, associated with color, the organism‟s morphology (i.e. weight, volume) is an important aspect that may be related to individual responses to solar radiation. Ultraviolet radiation causes serious mutagenic deleterious effects on organisms due to the high photon energy, causing mortality (21), development inhibition (22) and affecting species interactions (23,24). As a consequence, many insects are associated with environments with low ultraviolet rates (25). Ultraviolet radiation has a high penetration power, and its transmittance into the body should be a negatively selected strategy, as individuals would suffer internal damages. Therefore, behavioral (26) or physical (27) mechanisms to avoid ultraviolet radiation are frequent in animals. The elytra of Coleoptera, which are composed of an exocuticle and endocuticle structured by chitin fibers in a matrix with lipids and proteins, have many functions (28-30). The cuticle can act as barrier against external influences, such as water (31,32), pathogens (33), and radiation (34). In this study we examine, for the first time, the elytra spectrophotometric response of Neotropical dung beetles, distributed along an agricultural landscape, to the three aforementioned types of radiation. We used a spectrophotometer to characterize the transmittance, reflectance, and absorbance of the elytra to radiation with different wavelengths, in order to (I) derive common patterns, as well as (II) to examine the correspondence between these patterns and the species habitat preference along the agricultural landscape.

Materials and Methods

Origin of specimens Specimens come from a survey conducted in the region of São Miguel do Oeste, Santa Catarina (SC), southern Brazil (26°43'31''S, 53°31'05''W) in the summer of 2015. The surveyed area is an agricultural landscape composed mainly of maize crops with several patches of Atlantic Forest found within this agricultural matrix. Sampling was carried out at 30 sites - 15 agricultural crop areas and 15 Atlantic Forest remnants. For each site we placed five pitfall traps baited with human dung and five baited with decomposing pork meat, which were left in the field for 48 hours. All the collected insects were deposited in the entomological collection of the Federal University of Santa Catarina. Five Canthon species were selected for this study in 124

order to: minimize the occurrence of different responses due to phylogenetic differences, considering just diurnal species (20), and using a pool of species manifesting different environmental preferences (forest or maize crop areas). Thus, considering abundance data (Table S1 in supplementary material) general habitat preferences were established according to the percentage of individuals collected within forest and agricultural areas. Two species seemed to preferentially inhabit forest areas: C. angularis Harold, 1868 (99% of individuals) and C. lividus lividus Blanchard, 1845 (100%). Two other species preferentially inhabited open agricultural areas: C. chalybaeus Blanchard, 1845 (91%) and C. tetraodon Blanchard, 1845 (100%), while C. quinquemaculatus Castelnau, 1840 appears both in forest (75%) and agricultural areas (25%). These species varied in body weight from 0.020 to 0.054 g and presented a blue-green coloration, except for C. quinquemaculatus that had orange-yellow coloration with black areas (Fig. 1).

Body measurements A total of 45 specimens of the formerly mentioned species were analyzed (n = 10, 9, 10, 6 and 10 for C. angularis, C. lividus lividus, C. chalybaeus, C. tetraodon and C. quinquemaculatus, respectively). To evaluate whether reflectance, transmittance or absorbance rates are influenced by the morphological characteristics of the individuals all were weighed using a Tx423L Shimadzu® balance with a precision of ± 0.001g. Body thickness (BT), body width (BW ) and body length (BL) were measured for each individual using a stereoscopic microscope (Fig.

1) in order to calculate their body volume as: .

Elytra thickness was also measured in its inner central part, and the three morphological variables (body weight, body volume and elytra thickness) were used as continuous covariates.

125

Figure 1. Elytra color variation of the five dung beetle species used in the study: (A) C. angularis (B) C. lividus lividus (C) C. chalybaeus (D) C. tetraodon and (E) C. quinquemaculatus; and estimated dung beetle body measurements, the shaded region indicates the anatomical position of the elytra removed for analysis.

Spectrophotometric Analysis The left elytra of each individual was carefully removed with tweezers and placed in a Shimadzu® UV-2600 spectrophotometer to calculate both reflectance (R) and transmittance (T). Absorbance (A) is estimated as 100-(R+T) (13). This spectrophotometer continuously scans the elytra surface with different monochromatic wavelengths ranging from 185 to 1400 nm at 1 nm intervals. Three scans were performed for each elytra. The coefficient of variation (CV) between these three measurements is very low both for transmittance (CV = 1.5%) and for reflectance (CV = 0.3%) or absorbance (CV = 0.2%). Thus, average values of T, R and A are calculated taking into account these three repetitions. The complete electromagnetic spectrum examined was divided in three ranges: ultraviolet (UV) from 185 to 390 nm, visible (VIS) from 391 to 749 nm, and near infrared (NIF) from 750 to 1400 nm. Average T, R and A percentages are calculated for each one of the spectral ranges in all the individuals, and this percentage considered as the response variable in statistical analyses. In the specific case of NIF radiation T, R and A are calculated both for the dorsal and 126

ventral part of the elytra in order to examine the possible elytral capacity of reflecting or transmitting the heat generated by the beetle‟s body (i.e. to evaluate whether there are different responses when the radiation reaches the body from an external heat source and when it tries to leave the body). In order to examine the influence of the elytra placement in the spectrophotometer we changed this position three times in an individual of C. quinquemaculatus always obtaining the same pattern with CV values of 18% for transmittance, 6% for reflectance and 5% for absorbance. This influence was only examined in C. quinquemaculatus because the spectrophotometer ray reached almost entirely the elytra area of the other Canthon species.

Statistical analyses General Linear Models (GLMs) with a type III sum of squares (i.e., estimating the partial effects of each factor while controlling for the effects of the remaining predictors) were used to estimate the differences between species in the percentages of transmittance, reflectance, and absorbance for UV, VIS, and NIF spectral ranges. „Species‟ was included as a fixed factor in all the analyses, while elytra side (dorsal vs. ventral) was also a fixed factor in the case of NIF radiation. The three morphological variables (weight, volume and elytra thickness) were included as covariates following an ANCOVA model design. Type III sum of squares allows the calculation of the partial effect of the species factor while controlling for the effects of the morphological covariates (from now on named pure variability). The difference between this pure explained variability and the explanatory capacity of the species factor when included alone will be high when the two kinds of predictors covary. Since the morphological characteristics of the individuals also form part of the distinctive morphology of the species, this explanatory difference can be attributed to the joint or inextricable effect of all the considered predictors. Model residuals were checked for normality and homoscedasticity assumptions. All statistics were performed using StatSoft‟s STATISTICA v12.0 (35).

Results The species present a similar pattern across the complete examined wavelength spectrum (Fig. 2). At higher wavelengths (NIF radiation) transmittance and absorbance are the major elytra responses. However, as the wavelength decreases, absorbance becomes the main elytra response (Fig. 2). 127

Figure 2. Absorbance, reflectance and transmittance percentages across the complete spectrum for the five considered dung beetle species: (A) C. angularis, (B) C. lividus lividus, (C) C. chalybaeus, (D) C. tetraodon and (E) C. quinquemaculatus. Shaded areas represent variation in the values obtained for the different individuals. 128

On average, UV and VIS transmittances are low (less than 2% and 7%, respectively), while reflectance percentages are similar, albeit modest (≈ 3%). Thus, most part of the UV and VIS radiation would be absorbed by the elytral cuticle (around 96% in the case of UV and 90% for VIS). However, although NIF dorsal reflectance continues to be moderate (≈ 11%), transmittance and absorbance percentages are relatively high (around 46% and 43%; see Table 1).

Table 1. Mean values (± 95% confidence interval) of absorbance (a), transmittance (t) and reflectance (r) measurements under near infrared (NIF), visible (VIS) and ultraviolet (UV) radiation of the five considered species of dung beetles. C.lividus C. C. angularis C. chalybaeus C. tetraodon lividus quinquemaculatus UV (a) 96.13 ± 0.96 94.19 ± 2.325 95.76 ± 1.269 95.96 ± 1.58 97.02 ± 0.089 UV (t) 1.27 ± 1.33 3.01 ± 1.40 1.57 ± 1.33 1.28 ± 1.71 0.04 ± 1.33 UV (r) 2.59 ± 0.15 2.80 ± 0.16 2.67 ± 0.15 2.75 ± 0.15 2.93 ± 0.15 VIS (a) 94.02 ± 1.24 92.17 ± 2.65 94.36 ± 1.12 92.11±2.14 79.94 ± 2.479 VIS (t) 3.62 ± 1.89 5.37 ± 1.99 3.35 ± 1.89 5.20 ± 2.43 15.13 ± 1.89 VIS (r) 2.36 ± 0.23 2.46 ± 0.24 2.29 ± 0.23 2.69 ± 0.29 4.94 ± 0.23 NIF (a) 45.72 ± 2.16 45.169 ± 1.90 45.014±3.183 41.36 ± 3.53 26.940 ± 4.177 NIF (t) 46.08 ± 2.44 45.55 ± 2.575 46.52 ± 2.442 45.34 ± 3.15 53.70 ± 2.442 NIF (r) 8.19 ± 0.987 9.27 ± 1.040 8.45 ± 0.987 13.29 ± 1.27 19.35 ± 1.040

Ultraviolet (UV; 185-390 nm) and visible radiation (VIS; 391- 749 nm) Neither transmittance (F(7, 37)=1.84, p=0.11) nor reflectance (F(7, 37)=1.98, p=0.08), nor absorbance (F(7, 37)=1.74, p=0.13) in the UV spectrum can be significantly explained by the considered predictors. This implies that the low UV transmittance and reflectance values as well as the high UV absorbance percentages do not vary among the analyzed species. The considered morphological predictors (body weight, volume and elytra thickness, Table S2 in supplementary material) do not interfere with UV radiation responses. Therefore, different morphologies (as well as different species) have the same pattern of reflectance, transmittance and absorption of UV radiation. However, VIS transmittance, reflectance and absorbance variation are accounted for by the selected explanatory variables (R2 x 100 = 75.9%, 91.6% and 83.1%, respectively). VIS transmittance values significantly vary between the considered species (Table 2), which is a 129

factor able to explain around 12% of total variability independently of the considered morphological predictors (i.e. pure variability). Since the variability accounted for by the factor species when considered alone is 71%, most (≈ 83%) of the explained variability in the VIS transmittance would be due to the joint or inextricable effect of all the considered predictors. A post hoc test indicates that the only statistically significant difference between the species is that of C. quinquemaculatus with the remaining species (Fig. 3-A), where this species has higher transmittance values. Body volume barely explains 4.4% of total variability (higher transmittance at larger volumes). VIS reflectance percentages also significantly differ between the species (Table 2), where its pure explained variability is 15.1%. Again, the variability accounted for by the species factor alone is very high (91%), so that mainly the joint effects of the factor species associated with their morphological characteristics must explain VIS reflectance. Furthermore, post hoc tests show that C. quinquemaculatus is the unique species with significantly high and different reflectance values (Fig. 3- A). Lastly, VIS absorbance results show a similar pattern. The species factor (13.4% of total pure variability) and body volume (3.7%) are the only statistically significant predictors, where the individual explanatory capacity of the factor species is also very high (79%). The VIS absorbance values for C. quinquemaculatus are unique and significantly lower and different from those of other species (post hoc test).

Table 2. GLM results of transmittance, reflectance and absorbance values in the visible (VIS) wavelength spectrum (391-749 nm). Transmittance Reflectance Absorbance Variables F p F p F p d.f Species 4.75 0.003 16.49 <0.0001 7.32 0.0001 (4,37) Weight 3.08 0.09 0.86 0.36 3.65 0.06 (1,37) Volume 6.80 0.01 1.87 0.18 8.06 0.007 (1,37)

Elytra 0.64 0.43 0.32 0.57 0.78 0.38 (1,37) thickness

130

Figure 3. Absorbance (squares), transmittance (circles), reflectance (triangles) and adjusted mean values (± 95% CI) of the different species in the (A) visible electromagnetic spectrum and in the (B) near infrared electromagnetic spectrum when all the considered are held at their means (i.e. controlling for the effect of the three covariates included in the model; body weight, volume and elytra thickness).

131

Near infrared radiation (NIF; 750-1400 nm) Total NIF transmittance, reflectance and absorbance variation can be accounted for by the selected explanatory variables (R2 x 100 = 61.0%, 91.3% and 86.0%, respectively). NIF transmittance percentages vary between the considered species, the elytra side and the body weight (Table 3). However, since the interaction term “species * elytra side” is not statistically significant the effect of the elytra side on NIF transmittance does not seem to vary among the considered species. The factor species is able to explain around 10% of total variability in NIF transmittance values independently of the considered morphological predictors. Since the variability accounted for by the factor species is 25.8% when considered individually, more than half (≈ 60%) of the variability in NIF transmittance explained by the species factor would be due to the joint or inextricable effect of the considered predictors. Post hoc tests show that the only statistically significant differences between the species are those, which involve C. quinquemaculatus from the remaining species (Fig. 3 B), since this species has a higher percentage of NIF transmittance. The elytra side accounted for 4.8% of the total NIF transmittance variability independent of the other predictors, being slightly higher on the ventral side (48.7% ± 1.7%; mean ± 95% CI) than on the dorsal side (45.9% ± 1.7%). The pure explanatory variability accounted for by body weight is 3.9%. Since the three morphological variables are able to explain 20.5% of total variability in NIF transmittance, most of the explanatory capacity should be again assigned to the inseparable effect of the considered predictors.

Table 3. GLM results of transmittance, reflectance and absorbance values in the near infrared (NIF) wavelength spectrum (750-1400 nm). Transmittace Reflectance Absorbance Variables F p F p F p d.f Species 3.09 0.02 10.17 <0.0001 9.02 <0.0001 (4,77) Elytra 5.85 0.02 5.02 0.03 12.26 0.0008 (1,77) side Species * Elytra 0.61 0.65 0.16 0.95 0.77 0.55 (4,77) side Weight 4.76 0.03 4.14 0.04 10.01 0.002 (1,77) Volume 3.42 0.07 19.26 <0.0001 14.57 0.002 (1,77) 132

Elytra 0.42 0.52 2.72 0.10 1.91 0.17 (1,77) thickness

NIF reflectance also varies between the species, the elytra side, the body weight and the volume (Table 3). The pure effect of the factor species accounts for 8.8% of the total variability. Again, the individual explanatory capacity of the factor species when considered alone is 73%, and most of the explained variability is due to the combined effect of the predictors. A post hoc test indicates that the statistically significant differences are due to the comparative higher NIF reflectance values of C. tetraodon and C. quinquemaculatus (Fig. 3-B). Elytra side explains only 1% of the variability, where the reflectance is slightly higher in the ventral area (12.1% ± 0.7%) than in the dorsal area (11.0 % ± 0.7%). Body weight also has an almost negligible explanatory capacity (hardly 1%), while body volume is able to explain 4% of total variability (higher NIF reflectance when the volume increases). As a consequence of the former results, absorbance also differs between the considered species purely explaining 12.2% of total variability (57% when considered individually). Post hoc tests indicate that the low absorbance values of C. quinquemaculatus are significantly different from all the other species (Fig. 3-B). The elytra side accounts for 4% of total variability, while body weight and volume explain 4% and 5% of total variability.

Discussion This is the first study that evaluated the reflectance, absorbance and transmittance capacities of the elytra in Neotropical dung beetles. The obtained results indicate that the subtle morphological differences between the considered species have a relatively small explanatory capacity (≈4%) and that most of the differences between species in reflectance, absorbance and transmittance are accounted for by the combined effect of morphological and species identity variables. Thus, morphological characteristics of the individuals form part of the distinctive features of the species, and both types of variables exercise a combined effect in explaining the obtained patterns. Most importantly, the provided results also pinpoint that all the selected species show a relatively similar pattern: light of shorter wavelengths and higher frequencies is almost entirely absorbed, while radiation of longer wavelengths and lower frequencies can largely pass through the elytra. 133

Species did not show differences in the response to UV radiation. Thus, this highly energetic source of radiation is barely reflected and most of it is absorbed by the elytra (around 96%). Due to the serious damage that can be caused by the exposure to UV radiation, such as cell mutation (36), DNA damage (37) and metabolic stress (38), we expected that the elytra would reflect most of the UV radiation. Some insects have structures able to reflect UV radiation, such as Lepidoptera in which this radiation is reflected by ultrastructures found in the wings (39). In the case of Coleoptera, Pope and Hinton (27) reported UV reflectance in several families (including Scarabaeidae), but this reflectance occurs in very specific body regions covered with setae and secretions. Other authors have suggested that the surface wax of insect cuticles would serve to reflect UV (40), and that elytra reticulation patterns could also reflect UV radiation (41). UV reflectance can be a negatively selected characteristic in the case of diurnal insects, such as Canthon species (20). Although UV reflectance seem to be used for communication in Lepidoptera, diurnal species that reflect this radiation attract predators, mainly birds (42, 43). Birds can see ultraviolet light (44), and are predators of dung beetles (45). Thus, we suspect that the lack of UV reflectance in the studied species may be associated, at least partially, with protection against predation. UV radiation and shorter visible wavelengths are in large part absorbed by the elytra. The absorption of the energy coming from a light with a shorter wavelength may be followed by the emission of light with a longer wavelength in the process known as fluorescence. Vulinec (46) showed that some Scarabaeinae species of the genus Phanaeus have areas that fluoresce under UV light, probably for sexual purposes. Other Coleoptera also absorb UV light and fluoresce, such as Cteniopus sulphureus in which individuals reflect a strong yellow light (the complementary color of violet) when subjected to ultraviolet light (47). In any case, the elytra of the studied dung beetles of the genus Canthon do not reflect UV light or fluoresce when submitted both to a UV source of 380 nm and 254 nm (data not shown). What is the fate of this absorbed radiation? If absorbed energy does not generate fluorescence then a plausible possibility is that these photons excite the atoms in the elytra converting this energy into heat that may be further used in the thermoregulation process (48). Recent evidence shows that beetles reach higher internal body temperatures when submitted to artificial sunlight than under infrared radiation (49). However, it is necessary to design additional experiments to verify this supposition and to estimate the amount of body heat coming from the exposure to shorter wavelengths. 134

In the case of visible radiation, only C. quinquemaculatus seems to show a distinctive pattern associated with its distinctive coloration (see Fig. 2 and Figure S1). This species exhibits slightly higher transmittance and reflectance values than the other studied species, as well as higher reflectance values in the orange-red spectrum of visible light (from 590 to 750 nm). For the remaining species the small observed peaks appear in the range from 400 to 500 nm (Figure S1) according to the blue-green coloration of C. angularis, C. chalybaeus, C. lividus lividus, and C. tetraodon. The different coloration observed in C. quinquemaculatus may be attributed to mimicry of a Hymenoptera pattern, and this mimicry may be linked with the thermoregulatory capacity of the elytra. Batesian mimicry is widely observed in insects and several species mimic dangerous bees and wasps. Mimetic species usually present a yellowish or orange color interspersed with black regions, and frequently exhibit similar behavior. This mimicry has been described in Diptera (50), Coleoptera (51), Lepidoptera (52) and Neuroptera (53) but not in Scarabaeinae, although the coloration pattern of C. quinquemaculatus is convergent in several species of this genera, as well as in some species of other genera, including Deltochilum, Scybalocanthon and Canthidium. The reflective capacity of the elytra seems to be an ancestral feature described for a 50 million year fossil (54). In Coleoptera, reflectance patterns seem to be smaller under visible radiations (55), and only some beetles with metallic colors have high reflectance values (56). In our study reflectance values are modest reaching a maximum of 20% across the entire spectrum, including visible color peaks (reaching 5% of reflectance). Thus, the elytra of Canthon species seems to be opaque to UV and short visible wavelengths, but almost transparent to longer visible wavelengths and near infrared radiation. A similar pattern has also been described for three Lepidoptera species (58). Furthermore, previous studies have shown that the initial body warm up of dung beetle species would be passively facilitated by the permeability of the exoskeleton to infrared radiation (49). Consequently thermal radiations can move in and out of the body, and beetles may acquire or eject the heat relatively easy. However, NIF transmittance seems to be comparatively higher in C. quinquemaculatus, so that the cuticle of this species lets heat pass through much more so than other species. Interestingly, this elytral transparency to heat seems to be slightly higher from the inside independently of the species, suggesting that the elytral structure could 135

comparatively better facilitate the removal of body heat. Additional morphological studies should be carried out studying the structure of the elytra exoskeleton in order to find the possible characteristics that generate these differences, and estimating the distribution of NIF transmittance values among dung beetle species. Although provisional, there is not a clear correlation between the relative occurrence of the species in the two examined habitats and their elytral responses to the different radiation sources. Species with similar habitat preferences, such as C. angularis and C. lividus lividus in forest areas, and C. chalybaeus and C. tetraodon in agricultural areas seem to manifest similar spectrophotometric responses. Only C. quinquemaculatus, a species that can inhabit in different kinds of habitats, shows a different pattern: both infrared and visible radiations with longer wavelengths can permeate their elytra in both sides thus heating its body or allowing the dissipation of body heat. Thus, this high transmittance could be associated with a faster acquisition of heat in conditions of low solar radiation, such as those under wooded environments or at dusk and dawn, but also be advantageous to avoid overheating in open areas (with more solar radiation) through the expulsion of internal body heat. Besides thermal advantage, the presence of C. quinquemaculatus in open areas may be favored by their protection against predation due to its bee-like coloration pattern. In any case, more studies comprising a higher number of species with contrasting environmental preferences are necessary to determine the correlation between spectrophotometric responses and the ecological or biogeographical characteristics of species.

Acknowledgments This research was funded by the “PVE” project MEC/MCTI/CAPES/CNPq/FAPs No 09/2014 of the Ministry of Education of Brazil: “Efeito comparado do clima e das mudanças no uso do solo na distribuição espacial de um grupo de insetos indicadores (Coleoptera: Scarabaeinae) na Mata Atlântica”. VMA would like to thank CAPES (Ministry of Education of Brazil) for a PhD Grant during a research stay in the Museo de Ciencias Naturales de Madrid (Process 88887.122299/2016-00). MIMH would like to thank CNPq (Science and Technology Ministry of Brazil, Proc. 309030/2013-7) for a Productivity Grant.

136

Supplementary Materials

Table S1, S2 and Figure S1 can be found at DOI: 10.1562/2006- xxxxxx.s1.

References 1. Hessen, D.O. (2008) Solar radiation and the evolution of life. In: Solar Radiation and Human Health. (Edited by Espen Bjertness), pp. 123-134 The Norwegian Academy of Science and Letters, Olso. 2. Wild, M., H. Gilgen, A. Roesch, A. Ohmura, C.N. Long, E.G. Dutton, B. Forgan, A. Kallis, V. Russak and A. Tsvetkov (2005) From dimming to brightening: decadal changes in solar radiation at earth‟s surface. Science. 308, 847-850. 3. Taylor, F. (1981) Ecology and evolution of physiological time in insects. Am. Nat. 117, 1-23. 4. Murdock, C.C., K.P. Paaijmans, A.S. Bell, J.G. King, J.F. Hillyer, A.F. Read, and M.B. Thomas (2012) Complex effects of temperature on mosquito immune function. Proc. Royal Soc. B. 279, 3357–3366. 5. Yang, P., J.R. Carey, and R.V. Dowell (1994) Temperature influence on the development and demography of Bactrocera dorsalis (Diptera: Tephritidae) in China. Environ. Entomol. 23, 971– 974. 6. Jayatilaka, P., A. Narendra, S.F. Reid, P. Cooper and J. Zeil (2011) Different effects of temperature on foraging activity schedules in sympatric Myrmecia ants. J. Exp. Biol. 214, 2730- 2738. 7. Crill, W.D., R.B. Huey and G.W. Gilchrist (1996) Within- and between-generation effects of temperature on the morphology and physiology of Drosophila melanogaster. Evolution. 50, 1205-1218. 8. Verdú, J.R., L. Arellano and C. Numa (2006) Thermoregulation in endothermic dung beetles (Coleoptera: Scarabaeidae): Effect of body size and ecophysiological constraints in flight. J. Insect Physiol. 52, 854-860. 9. Wittkopp, P.J. and P. Beldade (2009) Development and evolution of insect pigmentation: Genetic mechanisms and the potential consequences of pleiotropy. Semin. Cell. Dev. Biol. 20, 65-71. 137

10. Katayama, N., J.K. Abbott, J. Kjaerandsen, Y. Takahashi and E.I. Svensson (2014) Sexual selection on wing interference patterns in Drosophila melanogaster. PNAS. 111, 15144-15148. 11. Schaefer, H.M. and N. Stobbe (2006) Disruptive coloration provides camouflage independent of background matching. Proc. Royal Soc. B. 273, 2427 – 2432. 12. Joron, M. (2009) Aposematic coloration. In Encyclopedia of Insects, 2nd Edition. (Edited by Vincent H. Resh and Ring T Cardé), pp. 33-38. Academic Press, New York. 13. Kinoshita, S. (2008) Structural colors in the realm of nature. World scientific. 14. Pye, J.D. (2010) The distribution of circularly polarized light reflection in the Scarabaeoidea (Coleoptera). Biol. J. Linn. Soc. 100, 585 – 596. 15. Akamine, M., Ishikawa, K., Maekawa, K. and M. Kon (2011) The physical mechanism of cuticular color in Phelotrupes auratus (Coleoptera, Geotrupidae). Entomol. Sci. 14, 291 – 296. 16. Pereboom, J.J.M. and J.C Biesmeijer (2003) Thermal constraints for stingless bee foragers: the importance of body size and coloration. Oecologia. 137, 42-50. 17. Key, K. H. L. and M.F Day (1954) A temperature-controlled physiological colour response in the grasshopper Kosciuscola tristis Sjost (Orthoptera: Acrididae). Aust. J. Zool. 2, 309-339. 18. Key, K. H. L. and M.F Day (1954) The physiological mechanism of colour change in the grasshopper Kosciuscola tristis Sjost (Orthoptera: Acrididae). Aust. J. Zool. 2, 340-363. 19. May, M.L. (1979) Insect Thermoregulation. Annu. Rev. Entomol. 24, 313-349. 20. Hernandez, M.I.M. (2002) The night and day of dung beetles (Coleoptera, Scarabaeidae) in the Serra do Japi, Brazil: elytra colour related to daily activity. Rev. Bras. entomol. 46, 597-600 21. Beresford, G.W., G. Selby and J.C. Moore (2013) Lethal and sub-lethal effects of UV-B radiation exposure on the collembolan Folsomia candida (Willem) in the laboratory. Pedobiologia. 56, 89–95 22. Faruki, S.I., D.R. Das and S. Khatun (2005) Effects of UV- Radiation on the larvae of the lesser mealworm, Alphitobius diaperinus (Panzer) (Coleoptera:Tenebrionidae) and their progeny. J. Biol. Sci. 5, 444-448. 138

23. Bothwell, M.L., D.M.J. Sherbot and C.M. Pollock (1994) Ecosystem response to solar ultraviolet – B radiation: influence of trophic-level interactions. Science. 265, 97-100. 24. Rousseaux, M.C., C.L. Ballaré, A.L. Scopel, P.S. Searles, and M.M. Caldwell (1998) Solar ultraviolet-B radiation affects plant-insect interactions in a natural ecosystem of Tierra del Fuego (southern Argentina). Oecologia. 116, 528-535. 25. Mazza, C. A., M.M. Izaguirre, J. Zavala, A.L. Scopel and C.L. Ballaré (2002) Insect perception of ambient ultraviolet-B radiation. Ecol. Lett. 5, 722–726. 26. Wharton, D.R.A. (1971) Ultraviolet repellent and lethal action on the American cockroach. J. Econ Entomol. 64, 252-255. 27. Pope, R.D and H.E. Hinton (1977) A preliminary survey of ultraviolet reflectance in beetles. Biol. J. Linnean Soc. 9, 331- 348. 28. Vincent, J.F.V. and U.G.K. Wegst (2004) Design and mechanical properties of insect cuticle. Arthropod Struct. Dev. 33, 187-199. 29. van de Kamp, T., A. Riedel and H. Greven (2016) Micromorphology of the elytral cuticle of beetles, with an emphasis on weevils (Coleoptera : Curculionoidea). Arthropod Struct. Dev. 45, 14-22. 30. van de Kamp, T. and H. Greven (2010) On the architecture of beetle elytra. Entomologie heute. 22, 191–204. 31. Wigglesworth, V.B. (1985) Sclerotin and lipid in the waterproofing of the insect cuticle. Tissue Cell. 17, 227-248. 32. Hadley, N. F. (1982) Cuticle ultrastructure with respect to the lipid waterproofing barrier. J. Exp. Zool. 222, 239-248. 33. Ortiz-Urquiza, A. and N.O Keyhani (2013) Action on the surface: entomopathogenic fungi versus the insect cuticle. Insects. 4, 357-375. 34. Buck, N. and T.V. Callaghan (1999) The direct and indirect effects of enhanced UV-B on the moth caterpillar Epirrita autumnata. Ecol. Bull. 47, 68–76. 35. STATISTICA v. 12, StatSoft Inc, Tulsa, Oklahoma, USA. 36. Negishi, T., F. Tanabe and H. Hayatsu (1992) The genotoxicity of UVA irradiation in Drosophila melanogaster and the synergistic action of 8-methoxypsoralen and UVA. Carcinogenesis. 13, 1433–1436. 37. Negishi, T., C. Nagaoka, H. Hayatsu, K. Suziki, T. Hara, M. Kubota, M. Watanabe and K. Hieda (2001) Somatic-cell 139

mutation induced by UVA and monochromatic UV radiation in repair-proficient and deficient Drosophila melanogaster. Photochem. Photobiol. 73, 493–498. 38. Meng J.Y., C.Y. Zhang, F. Zhu, X.P. Wang and C.L. Lei (2009) Ultraviolet light-induced oxidative stress: Effects on antioxidant response of Helicoverpa armigera adults. J. Insect Physiol. 55, 588-592. 39. Silberglied, R.E. (1979) Communication in the ultraviolet. Ann. Rev. Ecol. Syst. 10, 373-398. 40. Hadley, N. F (1985) The Adaptive Role of Lipids in Biological Systems. Wiley InterScience, New York. 41. Drotz, M.K., T. Brodin, A.N. Nilsson (2010) Multiple origins of elytral reticulation modifications in the west palearctic Agabus bipustulatus complex (Coleoptera, Dytiscidae). PLoS One, 5, e9034. 42. Lyytinen, A., R.V Alatalo, L. Lindström and J. Mappes (2001) Can ultraviolet cues function as aposematic signals? Behav. Ecol. 12, 65–70. 43. Lyytinem, A., L. Lindström and J. Mappes (2004) Ultraviolet reflection and predation risk in diurnal and nocturnal Lepidoptera. Behav. Ecol. 15, 982-987. 44. Bennett, A.T.D. and I.C Cuthill (1993) Ultraviolet vision in birds: what is its function? Vision Res. 34, 1471-1478. 45. Young, O.P. (2015) Predation on dung beetles (Coleoptera: Scarabaeidae): a literature review. T. Am. Entomol. Soc. 141, 111-115. 46. Vulinec, K. (1997) Iridescent dung beetles: a different angle. Fla. Entomol. 80, 132-140. 47. Crowson, R.A. (1981) The biology of the coleoptera. Academic Press, Michigan university. 48. Volynchik, S., M. Plotkin, D.J. Bergman and J.S. Ishay (2008) Hornet flight activity and its correlation with UVB radiation, temperature and relative humidity. Photochem. Photobiol. 84, 81-85. 49. Amore, V., M.I.M Hernandez, L.M. Carrascal and J.M Lobo (2017) Exoskeleton may influence the internal body temperatures of Neotropical dung beetles (Col. Scarabaeinae). PeerJ. 5, e3349. 50. Edmunds, M. and T. Reader (2014) Evidence for batesian mimicry in a polymorphic hoverfly. Evolution, 68: 827-839. 140

51. Linsley, E.G. (1959) Mimetic Form and Coloration in the Cerambycidae (Coleoptera). Ann. Entomol. Soc. Am. 52, 125– 131. 52. Robinson, G.S. and D.J. Carter (1989) The first vespiform tineid moth (Lepidoptera: Tineidae). Syst. Entomol. 14, 259- 273. 53. Ohl, M. (2004) A new wasp-mimicking species of the genus Euclimacia from Thailand (Neuroptera, Mantispidae). Denisia. 13, 193-196. 54. Parker, A.R. and D.R. McKenzie (2003) The cause of 50 million-year old colour. Proc. R. Soc. Lond. B 270, S151-S153. 55. Crook, J.D., J.A. Francese, K.E. Zylstra, I. Fraser, A.J. Sawyer, D.W. Bartels, D.R. Lance and V.C. Mastro (2009) Laboratory and field response of the emerald ash borer (Coleoptera:Buprestidae) to selected regions of the electromagnetic spectrum. Forest Entomol. 102, 2160-2169. 56. Campos-Fernández, C., D.E. Azofeifa, M. Hernández-Jimenéz, A. Ruiz-Ruiz, W.E, Vargas (2011) Visible light reflection spectra from cuticle layered materials. Opt. Mater. Express. 1, 85-100. 57. Kinoshita, S., S. Yoshioka, Y. Fujii, and N. Okamoto (2002) Photophysics of structural color in the Morpho butterflies. Forma. 17, 103-121. 141

Supporting Information

Table S1. Abundance values collected for the five considered species of dung beetles in the surveyed forest fragments (FF) and maize crops (MC) within an agricultural landscape located in the region of São Miguel do Oeste, SC, Brazil. Area C. angularis C. lividus lividus C. chalybaeus C. tetraodon C.quinquemaculatus FF 1 23 0 1 0 5 FF 2 10 14 7 0 10 FF 3 0 0 4 0 15 FF 4 2 0 0 0 30 FF 5 0 0 1 0 12 FF 6 39 2 2 0 1 FF 7 15 2 1 0 7 FF 8 1 0 0 0 3 FF 9 0 0 0 0 21 FF 10 1 0 5 0 3 FF 11 0 0 0 0 22 FF 12 0 0 0 0 18 FF 13 0 0 0 0 7 FF 14 0 0 0 0 17 FF 15 0 0 0 0 3 Total (FF) 91 18 21 0 174 MC 1 1 0 16 0 5 MC 2 0 0 88 0 3 MC 3 0 0 0 0 0 142

MC 4 0 0 9 0 4 MC 5 0 0 1 0 0 MC 6 0 0 35 0 1 MC 7 0 0 11 0 4 MC 8 0 0 13 9 3 MC 9 0 0 4 0 6 MC 10 0 0 4 0 3 MC 11 0 0 4 0 3 MC 12 0 0 1 0 2 MC 13 0 0 13 0 14 MC 14 0 0 8 0 7 MC 15 0 0 1 0 2 Total (MC) 1 0 208 9 57

143

Table S2. Mean values (± 95% confidence interval) of the considered morphological variables used as covariates for five dung beetle species.

C.angularis C. lividus lividus C. chalybaeus C. tetraodon C. quinquemaculatus Weight (g) 0.020 ± 0.004 0.032 ± 0.005 0.021 ± 0.003 0.032 ± 0.011 0.054 ± 0.006 Volume (mm3) 293.25 ± 33.475 368.254 ±51.481 221.683 ± 21.277 451.942 ± 46.07 603.487 ± 51.988 Elytra thickness (mm) 0.083 ± 0.004 0.075 ± 0.004 0.088 ± 0.004 0.083 ± 0.006 0.121 ± 0.013 144

Figure S1. Visible spectral reflectance of the elytra of the different studied species

145

CONCLUSÕES GERAIS

Nesta tese, uma visão baseada na teoria do nicho ecológico foi utilizada para entender a distribuição espacial em diferentes hábitats dos besouros escarabeíneos, e tanto a abordagem tradicional da diversidade como a abordagem morfométrica foram utilizadas no entendimento da estruturação das comunidades de escarabeíneos frente às modificações antrópicas dos hábitats. Demonstramos que o uso de transgênicos, bem como o uso de insumos (inseticidas, fungicidas e herbicidas), causam efeitos negativos nas comunidades, sendo que as comunidades das matas adjacentes aos cultivos transgênicos apresentaram menor riqueza de espécies (Artigo I). Essa menor riqueza evidencia um possível efeito do milho transgênico nos escarabeíneos. A redução do fitness foi inferida através de modificações da forma do corpo em indivíduos de C. quinquemaculatus que apresentaram modificações morfométricas quando presentes em áreas de cultivos transgênicos. Essas alterações morfológicas podem trazer consequências para a sua reprodução, fisiologia e consequentemente para a sobrevivência, sendo que em longo prazo as espécies de escarabaíneos podem apresentar extinção local nas áreas sob influência do milho transgênico (Artigo II). Quando avaliamos a comunidade de um ponto de vista morfométrico, notamos que a diversidade morfométrica se correlaciona positivamente com a riqueza de espécies, principalmente devido à diferenciação de nichos, ou seja, a morfologia pode ser utilizada como um indicativo da partilha de recursos, e a morfometria pode ser aplicada no entendimento das relações de segregação de nicho, e por consequência na estruturação das comunidades. As comunidades das matas apresentam menores buracos no hipervolume morfométrico, em relação às comunidades dos cultivos. Assim, os maiores buracos provam a extinção local de espécies decorrente da conversão das áreas de mata nativa em cultivos agrícolas (Artigo III). Foi observado que a espécie C. quinquemaculatus consegue estabelecer populações tanto nas áreas de mata quanto nos cultivos agrícolas. Esse maior nicho de C. quinquemaculatus está associado à maior permeabilidade térmica que essa espécie possui. A maior permeabilidade permite que os indivíduos expulsem, através dos élitros, o calor do corpo quando presentes em áreas sob intensa radiação solar (áreas agrícolas). Assim como podem expulsar mais facilmente o calor, os indivíduos também podem transmitir com maior facilidade o calor para dentro do corpo, se aquecendo quando estão nas áreas com menor 146

intensidade de radiação solar (Artigo IV). Além disso, demonstramos que os élitros dos escarabeíneos possuem uma seletividade frente às radiações: radiações menos energéticas são transmitidas para dentro do corpo e, à medida que a radiação se torna mais energética, a transmitância começa a ser diminuída e o élitro passa a bloquer a radiação. Portanto, demonstramos que, além de agir como um protetor contra choques mecânicos e perda de água, os élitros também agem como bloqueadores de radiações nocivas. A combinação das diferentes análises utilizadas nesse trabalho contribuiu para o entendimento da estruturação das comunidades desse grupo indicador como reposta à conversão agrícola e ao uso de transgênicos. Também demonstramos a eficiência da visão morfométrica no estudo das populações e das comunidades de escarabeíneos, e contribuímos no entendimento do efeito fisiológico na seleção de hábitat destes organismos.

147

REFERÊNCIAS

ABREU, L.; CANSI, E. & JURIATTI, C. 2007. Avaliação do rendimento sócio-econômico de variedades crioulas e híbridos comerciais de milho na microregião de Chapecó. Revista Brasileira de Agroecologia, v. 2, p. 1230-1233.

ALVAREZ, A. F.; ORTEGO, F. & CASTANERA, P. 2009. Bt maize fed-prey mediated effect on fitness and digestive physiology of the ground predator Poecilus cupreus L. (Coleoptera: Carabidae). Journal of Insect Physiology, v. 55, p. 143–149.

ANDERSSON, H.; TAGO, D. & TREICH, N. 2014. Pesticides and health: A review of evidence on health effects, valuation of risks, and benefit‐cost analysis. Preference Measurement in Health, v. 36, p. 1- 60.

ARAÚJO, P. M. & NASS, L. L. 2002. Caracterização e avaliação de populações de milho crioulo. Scientia Agrícola, v. 59, p. 589-593.

AXELSSON, E. P.; HJALTEN, J.; LEROY, C. J.; WHITHAM, T. G.; JULKUNEN-TIITTO, R. & WENNSTROM, A. 2011. Leaf litter from insect-resistant transgenic trees causes changes in aquatic insect community composition. Journal of Applied Ecology, v. 48, p. 1472– 1479.

BEDOUSSAC, L.; FAVILA, M. E & LÓPEZ, R. M. 2007. Defensive volatile secretions of two diplopod species attract the carrion ball roller scarab Canthon morsei (Coleoptera: Scarabaeidae). Chemoecology, v. 17, n. 3, p. 163–167.

BLONDER, B.; LAMANNA, C.; VIOLLE, C. & ENQUIST, B. J. 2014. The n-dimensional hypervolume. Global Ecology and Biogeography, v. 23, n. 5, p. 595–609.

BLONDER, B. 2016. Do hypervolumes have holes? American Naturalist, v. 187, n. 4, p. 93-105.

BOUCHARD, P.; BOUSQUET, Y.; DAVIES, A. E.; ALONSO- ZARAZAGA, M. A.; LAWRENCE, J. F.; LYAL, C. H. C.; NEWTON, A. F.; REID, C. A. M.; SCHIMITT, M.; ŚLIPIŃSKI, S. A. & SMITH, 148

A. B. T. 2011. Family-group names in Coleoptera (Insecta) Zookeys, v.88, p. 1-972.

CAMPOS, R. C. & HERNÁNDEZ, M. I. M. 2015a. Changes in the dynamics of functional groups in communities of dung beetles in Atlantic forest fragments adjacent to transgenic maize crops. Ecological Indicators, v. 49, p. 216-227.

CAMPOS, R. C. & HERNÁNDEZ, M. I. M. 2015b. The importance of maize management on dung beetle communities in Atlantic Forest fragments. PLoS ONE, v. 6, e0145000.

CAMPOS, R. C. 2016. Influência do cultivo de milho transgênico em organismos não-alvo (Coleoptera: Scarabaeinae) e da ingestão indireta através da cadeia trófica. Tese de doutorado, Universidade Federal de Santa Catarina, Florianópolis.

CARPENTIERI-PÍPOLO, V.; SOUZA, A.; SILVA, D. A.; BARRETO, T. P.; GARBUGLIO, D. D. & FERREIRA, J. M. 2010. Avaliação de cultivares de milho crioulo em sistema de baixo nível tecnológico. Acta Scientiarum Agronomy, v. 32, p. 229-233.

CASARI, S. A. & IDE, S. 2012. Coleoptera, p. 454-552. In: RAFAEL, J. A.; MELO, G. A. R.; CARVALHO, C. J. B. & CONSTANTINO, R. (eds.). Insetos do Brasil Diversidade e Taxonomia. Holos Editora. 796 p.

CATÃO, H. C. R. M., MAGALHÃES, H. M., SALES, N. L. P., BRANDÃO, D. & ROCHA, F. S. 2013. Incidência e viabilidade de sementes crioulas de milho naturalmente infestadas com fungos em pré e pós-armazenamento. Ciência Rural, v. 43, p. 764-770.

CHASE, J. M. & LEIBOLD, M. A. 2001. Ecological Niches: Linking Classical and Contemporary Approaches. Chicago: University of Chicago Press. 212 p.

CHAPIN, F. E.; ZAVALETA, E. S.; EVINER, T. V.; NAYLOR, R. L.; VIOUSEK, P. M.; REYNOLDS, H. L.; HOOPER, D. U.; LAVOREL, S.; SALA, O. E.; HOBBIE, S. E.; MACK, M. C. & DÍAZ, S. 2000. Consequences of changing biodiversity. Nature, v. 405, p. 234-242.

149

COHN, A. S.; VANWEY, L. K.; SPERA, S. A. & MUSTARD, J. F. 2016. Cropping frequency and area response to climate variability can exceed yield response. Nature Climate Change, v. 6, p. 601–604.

COHEN, J. E. 1977. Food webs and the dimensionality of trophic niche space. Proceedings of the National Academy of Sciences USA, v. 74, n. 10, p. 4533 -4536.

CONNELL, J. H. 1980. Diversity and the coevolution of competitors, or the ghost of competition past. Oikos, v. 35, n. 2, p. 131-138.

CORNWELL, W. K.; SCHWILK, D. W. & ACKERLY, D. D. 2006. A trait-based test for habitat filtering: convex hull volume. Ecology, v. 87, n. 6, p. 1465–1471.

DOUGLAS, M. E. & MATTHEWS, W. J. 1992. Does morphology predict ecology? Hypothesis testing within a freshwater stream fish assemblage. Oikos, v. 65, n. 2, p. 213-224.

EPAGRI, EMPRESA DE PESQUISA AGROPECUÁRIA E EXTENSÃO RURAL DE SANTA CATARINA. 2015. Números da Agropecuária Catarinense. Disponível em: http://www.epagri.sc.gov.br (acessado em 10/11/2017).

EPAGRI, EMPRESA DE PESQUISA AGROPECUÁRIA E EXTENSÃO RURAL DE SANTA CATARINA. 2017. Boletim Agropecuário n º 47. Disponível em: http://www.epagri.sc.gov.br (acessado em 10/11/2017).

ESTRADA, A. & COATES-ESTRADA, R. 1991. Howler monkeys (Alouatta palliata), dung beetles (Scarabaeidae) and seed dispersal: ecological interactions in the tropical rain forest of Los Tuxtlas. Journal of Tropical Ecology, v. 7, p. 459-474.

ELFERINK, M. & SCHIERHORN F. 2016. Global demand for food is rising. Can we meet it? Harvard Business Review. Disponível em: www.hbr.og. (acessado em 15/10/2017).

150

FAO, FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED STATES (FAO). 2015. How to feed world in 2050. Disponível em: www.fao.org. (acessado em: 15/09/2017)

FAO, FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED STATES. 2017. Food and agriculture data. Disponível em: http://www.fao.org./faostat (acessado em: 15/09/2017)

FARIAS, P. M. & HERNÁNDEZ, M. I. M. 2017. Dung beetles associated with agroecosystems of southern Brazil: relationship with soil properties. Revista Brasileira de Ciência do Solo, v. 41, e0160248.

FINCHER, G. 1981. The potential value of dung beetles in pasture ecosystems. Journal of the Georgia Entomological Society, v. 16, p. 316-333.

FORSMAN, A. & SHINE, R. 1995. Parallel geographic variation in body shape and reproductive life history within the Australian scincid lizard Lampropholis delicata. Functional Ecology, v. 9, n. 6, p. 818- 828.

GALBIATI, C., BENSI, C., CONCCICAO, C. H. C., FLORCOVSKI, J. L., CALAFI, M. H. & TOBIAS, A. C. T. (1995) Estudo comparativo entre besouros do esterco Dichotomius analypticus (Mann, 1829). Ecossistema, v. 20, p. 109-118.

GARDNER, T. A.; HERNÁNDEZ M. I. M.; BARLOW, J. & PERES, C. A. 2008a. Understanding the biodiversity consequences of habitat change: the value of secondary and plantation forests for neotropical dung beetles. Journal of Applied Ecology, v. 45, p. 883-893.

GARDNER, T. A.; BARLOW, J.; ARAUJO, L. S.; ÁVILA-PIRES, T. C.; BONALDO, A. B.; COSTA, J. E.; ESPOSITO, M. C.; FERREIRA, L. V.; HAWES, J.; HERNÁNDEZ, M. I. M.; HOOGMOED, M. S.; LEITE, R. N.; LO-MAN-HUNG, N. F.; MALCOLM, J. R.; MARTINS, M. B.; MESTRE, L. A. M.; MIRANDA-SANTOS, R.; NUNESGUTJAHR, A. L.; OVERAL, W. L.; PARRY, L.; PETERS, S. L.; RIBEIRO-JUNIOR, M. A.; SILVA, M. N. F.; SILVA-MOTTA, C. & PERES, C. A. 2008b. The cost-effectiveness of biodiversity surveys in Tropical Forests. Ecology Letters, v. 11, n. 2, p. 139-150.

151

GASSMANN, A. J.; PETZOLD-MAXWELL J. L.; KEWESHAN, R.S. & DUNBAR M. W. 2011. Field-evolved resistance to Bt maize by western corn rootworm. PLoS ONE, v. 6, e22629.

GASTON, K. A.; EFT, J. A. & LAUER, T. E. 2012. Morphology and its effect on habitat selection of stream fishes. Proceedings of the Indiana Academy of Science, v. 121, p. 71-78.

GAUSE, G. F. 1936. The struggle for existence. Baltimore: Williams and Wilkins, 200p.

GILLARD, P. 1967. Coprophagus beetles in pasture ecosystems. Journal of Australian Institute for Agricultural Sciences, v. 33, p. 30-34.

GRINNEL, J. 1917. The niche-relationships of the California thrasher. The Auk, v. 34, n. 4, p. 427 – 433.

HALFFTER, G. & MATTHEWS, E. G. 1966. The natural history of dung beetles of the subfamily Scarabaeinae (Coleoptera: Scarabaeidae). Folia Entomológica Mexicana, v. 12-14, p. 1-312.

HALFFTER, G. & EDMONDS, W. D. 1982. The nesting behavior of dung beetles (Scarabaeinae): An ecologic and evolutive approach. México D.F.: Man and Biosphere Program UNESCO, 177 p.

HALFFTER, G. 1991. Historical and ecological factors determining the geographical distribution of beetles (Coleoptera: Scarabaeidae: Scarabaeinae). Folia Entomológica Mexicana, v. 82, p. 195-238.

HALFFTER, G. & FAVILA, M. E. 1993. The Scarabaeidae (Insecta: Coleoptera) an animal group for analyzing, inventorying and monitoring biodiversity in tropical rainforest and modified landscapes. Biology International, v. 27, p. 15-21.

HAN, P.; NIU, C. Y.; LEI, C. L. & DESNEUX, C. 2010. Use of an innovative T-tube maze assay and the proboscis extension response assay to assess sublethal effects of GM products and pesticides on learning capacity of the honey bee Apis mellifera L. Ecotoxicology, v. 19, n. 8, p. 1612-1619.

152

HANSKI, I. & CAMBEFORT, Y. 1991. Dung beetle ecology. Princeton: Princeton University Press. 481 p.

HENDRICKX, F.; MAELFAIT, J. P.; WINGERDEN, W.; SCHWEIGER, O.; SPEELMANS, M.; AVIRON, S.; AUGESTEIN, I.; BILLETER, R.; BAILEY, D.; BUKACEK, R.; BUREL, F.; DIEKOTTER, T.; DIRKSEN, J.; HERZOG, F.; LIIRA, J.; ROUBALOVA, M.; VANDOMME, V. & BUGTER, R. 2007. How landscape structure, land-use intensity and habitat diversity affect components of total arthropod diversity in agricultural landscapes. Journal of Applied Ecology, v. 44, n. 2, p. 340-351.

HERNÁNDEZ, M. I. M. 2002. The night and day of dung beetles (Coleoptera, Scarabaeidae) in the Serra do Japi, Brazil: elytra colour related to daily activity. Revista Brasileira de Entomologia, v. 46, n. 4, p. 597-600.

HERNÁNDEZ, M. I. M.; MONTEIRO, L. R. & FAVILA, M. E. 2011. The role of body size and shape in understanding competitive interactions within a community of Neotropical dung beetles. Journal of Insect Science, v. 11, n. 13, p. 1-14.

HOLZSCHUH, A.; STEFFAN-DEWENTER, I. & TSCHARNTKE, T. 2008. Agricultural landscapes with organic crops support higher pollinator diversity. Oikos, v. 117, p. 354-361.

HOFMANN, C.; VANDERBRUGGEN, H.; HOFTE, H.; VAN RIE, J.; JANSENS, S. & VAN MELLAERT, H. 1988. Specificity of Bacillus thuringiensis endotoxins is correlated with the presence of high-affinity binding sites in the brush border membrane of target insect midguts. Proceedings of the National Academy of Sciences, v. 85, p. 7844- 7848.

HUTCHINSON, G. 1957. Concluding remarks. Cold Spring Harbor Symposium on Quantitative Biology, v. 22, p. 415–427.

HUTTON, J. B.; JURY, K. E. & DAVIES E. B. 1967. Studies of the nutritive value of New Zealand dairy pastures. The intake and utilization of potassium, sodium, calcium, phosphorus and nitrogen in pasture herbage by lactating dairy cattle. New Zealand Journal of Agricultural Research, v. 10, p. 367-388. 153

IBGE, INSTITUTO BRASILEIRO DE GEOGRAFIA E ESTATÍSTICA. 2017. Censo Agropecuário. Disponível em: http:// www.ibge.gov.br (acessado em 05/07/2017).

INGHAM, E. R. 1985. Review of the effects of 12 selected biocides on target and non-target soil organisms. Crop Protection, v. 4, n.1, p. 3-32.

INGLESFIELD, C. 1989. Pyrethroids and Terrestrial Non-target Organisms. Pesticide Science, v. 27, p. 387-428.

INOUYE, D. W. 1980. The effect of proboscis and corolla tube lengths on patterns and rates of flower visitation by bumblebees. Oecologia, v. 45, n. 2, p. 197-201.

ISAAA. 2016. Global status of commercialized Biotech/GM Crops: 2016. Brief 49-2014: Executive Summary. Disponível em: http://isaaa.org (acessado em 10/04/2016)

JAMES, C. 2010. Global Status of Commercialized Biotech/GM crops. Brief No. 44, ISAAA, Ithaca, New York.

KEAST, A. & WEBB, D. 1966. Mouth and body form relative to feeding ecology in the fish fauna of a small lake, lake Opinicon, Ontario. Journal of Fisheries Research Board of Canada, v. 23, n. 12, p. 1845-1874.

KIMATU, J. N. 2011. Effects of narrowing genetic base and abiotic stress on leaf-spotting in Grevillea robusta. Journal of Tropical Forest Science, v. 23, n. 2, p. 117-124.

KLEIN, B. C. 1989. Effects of forest fragmentation on dung and carrion beetle communities in central Amazonia. Ecology, v. 70, n. 6, p. 1715- 1725.

KRAUSS, J.; STEFFAN-DEWENTER, I.; MULLER, C. B. & TSCHARNTKE, T. 2005. Relative importance of resource quantity, isolation and habitat quality for landscape distribution of a monophagous butterfly. Ecography, v.28, n.4, p. 465-474.

154

MACARTHUR, R. H. & LEVINS, R. 1964. Competition, habitat selection, and character displacement in a patchy environment. Proceedings of the National Academy of Sciences USA, v. 51, n. 6, p. 1207 – 1210.

MACARTHUR, R. H. 1958. Population ecology of some warblers of northeastern coniferous forests. Ecology, v. 39, n. 4, p. 599–619.

MACARTHUR, R. H. & LEVINS, R. 1967. Limiting similarity, convergence and divergence in coexisting species. American Naturalist, v. 101, n. 921, p. 377–385.

MARVIER, M.; MCCREEDY, C.; REGETZ, J. & KAREIVA, P. 2007. A meta-analysis of effects of Bt cotton and maize on non-target invertebrates. Science, v. 8, n. 316, p. 1475–1477.

MATSON, P.A.; PARTON, W. J.; POWER, A. G. & SWIFT, M. J. 1977. Agricultural intensification and ecosystem properties. Science, v. 277, n. 5352, p. 504-509.

MAY, R. M. 1973. Stability and complexity in model ecosystems. Princeton, Princeton University Press, 304 p.

MCCAULEY, S. J.; DAVIS, C. J.; RELYEA, R. A.; YUREWICZ, K. L.; SKELLY, D. K.; WERNER, E. E. 2008. Metacommunity patterns in larval odonates. Oecologia, v. 158, n.2, p. 329-342

MCCAULEY, S. J.; DAVIS, C. J.; WERNER, E. E. & ROBENSON, M. S. 2014. Dispersal, niche breadth and population extinction: colonization ratios predict range size in North American dragonflies. Journal of Animal Ecology, v. 83, n.4, p. 858-865.

MICHAUX, J.; CHEVRET, P. & RENAUD, S. 2007. Morphological diversity of Old World rats and mice (Rodentia, Muridae) mandible in relation with phylogeny and adaptation. Journal of Zoological Systematics and Evolutionary Research, v. 45, n.3, p. 263-279.

MORSE, N. B.; PELLISSIER, P. A.; CIANCIOLA, E. N.; BRERETON, R. L.; SULIVAN, M. M.; SHONKA, N. K.; WHEELER, T. B. & MCDOWELL, W. H. 2014. Novel ecosystems in the 155

Anthropocene: a revision of the novel ecosystem concept for pragmatic applications. Ecology and Society, v. 19, n. 2: 12.

MURCIA, C.; ARONSON, J.; KATTAN, G. H.; MORENO-MATEOS, D.; DIXON, K. & SIMBERLOFF, D. 2014. A critique of the “novel ecosystem” concept. Trends in Ecology and Evolution, v. 29, n.10, p. 548-553.

NERVO, B.; TOCCO, C.; CAPRIO, E.; PALESTRINI, C. & ROLANDO, A. 2014. The effects of body mass on dung removal efficiency in dung beetles. PLoS ONE, v. 9, e107699.

NICHOLS, E., LARSEN, T., SPECTOR, S., DAVIS, A. L., ESCOBAR, F., FAVILA, M. & VULINEC, K. 2007. Global dung beetle response to tropical forest modification and fragmentation: A quantitative literature review and meta-analysis. Biological Conservation, v. 137, p. 1-19.

NICHOLS, E.; SPECTOR, S.; LOUZADA, J.; LARSEN, T.; AMÉZQUITA, S. & FAVILA, M. E. 2008. Ecological functions and ecosystem services provided by Scarabaeinae dung beetles. Biological Conservation, v. 141, p. 1461-1474.

NOBRE, C. A.; SAMPAIO, G.; BORMA, L. S.; CASTILLA-RUBIO, J. C.; SILVA, J. S. & CARDOSO, M. 2016. Land-use and climate change risks in the Amazon and the need of a novel sustainable development paradigm. Proceedings of the National Academy of Sciences of the United States of America, v. 113, n. 39, p. 10759–10768

ODUM, E. P. 1983. Fundamentos em Ecologia. Lisboa: Fundação Calouste Gulbekian. 927 p.

PATERNIANI, E., NASS, E. & SANTOS, M. X. 2000. O valor dos recursos genéticos de milho para o Brasil – uma abordagem histórica da utilização do germoplasma. In: Udry, C.V & Duarte, W. (eds.) Uma história brasileira do milho, o valor dos recursos genéticos. Brasília, Paralelo Brasília.

PETERS, R. H. 1983. The ecological implications of body size. Cambridge: Cambridge University Press, p. 344.

156

PIANKA, E. R. 1973. The structure of lizard communities. Annual Review of Ecology, Evolution, and Systematics, v. 4, p. 53-74.

PULLIAM, H. R. 2000. On the relationship between niche and distribution. Ecology Letters, v. 3, n. 4, p. 349 – 361.

RICKLEFS, R. E. & TRAVIS, J. 1980. A morphological approach to the study of avian community organization. The Auk, v. 97, n.2, p. 321–338.

RICKLEFS, R. E. 1996. A economia da natureza. Rio de Janeiro: Guanabara Koogan, 470 p.

ROOT, R. B. 1967. The niche exploitation pattern of the blue-gray gnatcatcher. Ecologial Monographs, v. 37, n. 4, p. 317 – 350.

ROWE, L. & ARNQVIST, G. 2012. Sexual selection and the evolution of genital shape and complexity in water striders. Evolution, v. 66, n. 1, p. 40-54.

SCHWEIGER, O.; MAELFAIT, J. P.; WINGERDEN, W.; HENDRICX, F.; BILLETER, R.; SPEELMANS, M.; AUGESTEIN, I.; AUKEMA, B.; AVIRON, S.; BAILEY, D.; BUKACEK, R.; BUREL, F.; DIEKOTTER, T.; DIRKSEN, J.; FRENZEL, M.; HERZOG, F.; LIIRA, J.; ROUBALOVA, M. & BUGTER, R. 2005. Quantifying the impact of environmental factors on arthropod communities in agricultural landscapes across organizational levels and spatial scales. Journal of Applied Ecology, v. 42, n.6, p. 1129-1139.

SIMMONS, L. W. & RIDSDILL-SMITH, T. J. 2011. Ecology and evolution of dung beetles. Oxford: Blackwell Publishing, 347 p.

SLADEN, W. J. L.; MENZIE, C. M. & REICHEL, W. L. 1966. DDT Residues in adelie penguin and a crabeater seal from Antarctica. Nature, v. 210, p. 670-673.

SNAPE, I.; SCOULLER, R. C.; STARK, S. C.; RIDDLE, M. J. & GORE, D. B. 1966. Characterisation of the dilute HCl extraction method for the identification of metal contamination in Antarctic marine sediments. Chemosphere, v. 57, n.6, p. 491-504.

157

SOBERÓN, J. & PETERSON, A. T. 2005. Interpretation of models of fundamental ecological niches and species‟ distributional areas. Biodiversity Informatics, v. 2, p. 1 – 10.

SOBERÓN, J. 2007. Grinnellian and Eltonian niches and geographic distributions of species. Ecology Letters, v. 10, n. 12, p. 1115–1123.

SOBERÓN, J. & NAKAMURA, M. 2009. Niches and distributional areas: Concepts, methods, and assumptions. Proceedings of the National Academy of Sciences, v. 106, n. 2, p. 19644 – 19650.

SOBERÓN, J. & ARROYO-PEÑA, B. 2017. Are fundamental niches larger than the realized? Testing a 50-year-old prediction by Hutchinson. PLoS ONE, v. 12, n. 4, e0175138.

SPENCER, L. M. 1995. Morphological correlates of dietary resource partitioning in the african bovidae. Journal of Mammalogy, v. 76, n. 2, p. 448-471.

TARASOV, S. & GÉNIER, F. 2015. Innovative bayesian and parsimony phylogeny of dung beetles (Coleoptera, Scarabaeidae, Scarabaeinae) enhanced by ontology-based partitioning of morphological characters. PLoS ONE, v.10, n.3, e0116671.

TILMAN, D. 1982. Resource competition and community structure. Princeton: Princeton University Press, 296 p.

TILMAN, D. 2004. Niche tradeoffs, neutrality, and community structure: a stochastic theory of resource competition, invasion, and community assembly. Proceedings of the National Academy of Sciences USA, v. 101, n. 30, p. 10854-10861.

TIN, T.; FLEMING, Z. L.; HUGHES, K. A.; AINLEY, D. G.; CONVEY, O.; MORENO, C. A.; PFEIFFER, S.; SCOTT, J. & SNAPE, I. 2008. Impacts of local human activities on Antarctic environment. Antarctic Science, v. 21, n. 1, p. 3-33.

TRIPLEHORN, C. A. & JOHNSON, N. F. 2011. Estudo dos insetos. São Paulo: Cengage Learning, 809 p.

158

VAZ-DE-MELLO, F.Z. 2018. Scarabaeinae in Catálogo Taxonômico da Fauna do Brasil. . (acessado em: 20/01/2018).

VOGT, G. A.; CANCI, I. J. & CANCI, A. 2007. Uso e manejo de variedades locais de milho em Anchieta (SC). Agriculturas, v.4, n.3, p. 36-39.

WILSON, J. D.; MORRIS, A. J.; ARRYOO, B. E.; CLARK, S. C. & BRADBURY, R. B. 1999. A review of the abundance and diversity of invertebrate and plant foods of granivorous birds in northern Europe in relation to agricultural change. Agriculture, Ecosystems and Environment, v. 75, p. 13-30.

WINANS, G. A. & NISHIOKA, R. S. 1987. A multivariate description of change in body shape of coho salmon (Oncorhynchus kisutch) during smoltification. Aquaculture, v. 66, p. 235-245.

WEST, K.; COHEN, A. & BARON, M. 1991. Morphology and behavior of crabs and gastropods from lake tanganyika, africa: implications for lacustrine predator-prey coevolution. Evolution, v. 45, n. 3, p. 189-607.