UNIVERSIDADE ESTADUAL DE CAMPINAS
Instituto de Biologia
Luis Francisco Prado Pinheiro Ferreira Salles
Efeitos de borda sobre artrópodes epigeicos no Cerrado
Edge effects on epigaeic arthropods in Cerrado
Campinas
2017
EFEITOS DE BORDA SOBRE ARTRÓPODES EPIGEICOS NO CERRADO
EDGE EFFECTS ON EPIGAEIC ARTHROPODS IN CERRADO
Dissertação apresentada ao Instituto de Biologia da Universidade Estadual de Campinas como parte dos requisitos exigidos para a obtenção do título de Mestre em Ecologia
Dissertation presented to the Institute of Biology of the University of Campinas in partial fulfillment of the requirements for the degree of Master in the area of Ecology
ORIENTADOR: Prof. Dr. Paulo Sérgio Moreira Carvalho de Oliveira
ESTE EXEMPLAR CORRESPONDE À VERSÃO FINAL DA DISSERTAÇÃO DEFENDIDA PELO ALUNO LUIS FRANCISCO PRADO PINHEIRO FERREIRA SALLES E ORIENTADA PELO PROF. DR. PAULO SÉRGIO MOREIRA CARVALHO DE OLIVEIRA
CAMPINAS 2017
Ficha catalográfica
Universidade Estadual de Campinas Biblioteca do Instituto de Biologia Gustavo Lebre de Marco - CRB 8/7977
Salles, Luis Francisco Prado Pinheiro Ferreira,
Sa34e SalEdge effects on epigaeic arthropods in Cerrado / Luis Francisco Prado Pinheiro Ferreira Salles. – Campinas, SP : [s.n.], 2017.
SalOrientador: Paulo Sergio Moreira Carvalho de Oliveira.
SalDissertação (mestrado) – Universidade Estadual de Campinas, Instituto de Biologia.
Sal1. Fragmentação territorial. 2. Formiga - Ecologia. 3. Grupos funcionais. 4. Ecologia de comunidades. I. Oliveira, Paulo Sérgio Moreira Carvalho de,1957-. II. Universidade Estadual de Campinas. Instituto de Biologia. III. Título.
Informações para Biblioteca Digital
Título em outro idioma: Efeitos de borda sobre artrópodes epigeicos no Cerrado Palavras-chave em inglês:
Territorial fragmentation Ants - Ecology
Functional groups Community ecology
Área de concentração: Ecologia Titulação: Mestre em Ecologia Banca examinadora:
Paulo Sérgio Moreira Carvalho de Oliveira André Victor Lucci Freitas
Tadeu de Siqueira Barros
Data de defesa: 10-07-2017
Programa de Pós-Graduação: Ecologia
COMISSÃO EXAMINADORA
Prof. Dr. Paulo Sérgio Moreira Carvalho de Oliveira (Orientador) Prof. Dr. André Victor Lucci Freitas
Prof. Dr. Tadeu de Siqueira Barros
Os membros da Comissão Examinadora acima assinaram a Ata de Defesa, que se encontra no processo de vida acadêmica do aluno.
“To become authentic is to passionately embrace your sense of self despite living in a world
that speciously lures its inhabitants away from subjectivity and honest introspection (...)”
Philip Lindholm
“There is a pleasure in the pathless woods,
There is a rapture on the lonely shore, There is society where none intrudes, By the deep Sea, and music in its roar: I love not Man the less, but Nature more (...)”
mas também por depositar sua confiança em mim ao aceitar me orientar como aluno de mestrado. Ele me mostrou a importância e a beleza do trabalho de um naturalista.
Agradeço ao Prof. Alexander Christianini, por me acompanhar desde a graduação e me ensinar tanto sobre Ecologia e sobre a profissão de pesquisador.
Agradeço aos membros titulares da banca, Prof. André Freitas, Prof. Tadeu Siqueira, e também aos membros suplentes da banca, Dr. Eduardo Barbosa e Profa. Luciane Junqueira. Tenho certeza que suas contribuições serão valiosas a este trabalho.
Agradeço aos membros da pré-banca, Prof. Alexander Christianini, Profa. Inara Leal e Dra. Paula Omena, que disponibilizaram seu tempo para ler este trabalho e fizeram sugestões que melhoraram consideravelmente este estudo.
Agradeço à UNICAMP, instituição que admiro e me orgulho por fazer parte, por fornecer a oportunidade a mim e a muitas pessoas de trabalhar e aprender em tantas áreas diferentes.
Agradeço ao Programa de Pós-Graduação em Ecologia pela oportunidade de me formar como Mestre em Ecologia por um programa tão renomado, que há tantos anos forma grandes ecólogos e professores de Ecologia.
Agradeço às equipes da Reserva Biológica de Mogi-Guaçu e da Estação Ecológica e Experimental de Itirapina, por me fornecerem permissão e infra-estrutura para realizar este estudo. Agradeço também por permanecerem se esforçando em cuidar de nosso Cerrado, esse bioma maravilhoso tão desvalorizado pela nossa sociedade.
Agradeço ao Prof. Rodrigo Feitosa e sua equipe, por me ajudarem a identificar minhas formigas e por terem me recebido muito bem em seu laboratório.
Agradeço à minha namorada e melhor amiga, Ellen, a quem tanto devo, por acreditar em mim quando eu mesmo não acredito e me trazer a felicidade que eu tenho certeza que não sentiria sem ela.
Agradeço aos meus amigos de laboratório: Javier, Mary, Maru, Hélio, Sebá, Nádia, André e Verônica, por toda a ajuda em campo, todo suporte, todas as conversas, risadas,
Agradeço às minhas amigas da UFSCar: Fran, Dani, Luiza, Larissa e Sarah. Sou muito grato a elas por permanecerem minhas amigas depois de terminada nossa graduação.
Agradeço, acima de tudo, aos meus pais, Denise e Adão, por me ensinarem o valor do conhecimento e da ética, e por me fornecerem a oportunidade de seguir a profissão que mais admiro, a de Biólogo. Não tenho a menor dúvida de que, se conquistei algo na vida, foi somente graças a eles.
Por fim, mas não menos importante, agradeço às formigas, esses pequenos seres que admiro imensamente e que me ensinam tanto sobre o que é a vida.
Este trabalho não poderia ter sido feito sem o apoio financeiro do CNPq (processo: 131488/2015-5) e da CAPES. Agradeço, também, à CAPES, pela concessão de bolsa PED durante o 1° semestre de 2017.
e uma de suas consequências são efeitos de borda. Efeitos de borda ocorrem em áreas de transição abrupta entre ecossistemas adjacentes e podem ter diversos efeitos negativos sobre processos ecológicos. O aumento no uso da terra para atividades humanas têm levado a savana brasileira, o Cerrado, a uma paisagem cada vez mais fragmentada, com um aumento esperado nas áreas remanescentes sujeitas a efeitos de borda, com consequentes perturbações nos papeis ecossistêmicos desempenhados por artrópodes. Este estudo avaliou efeitos de borda em artrópodes epigeicos, folhiço e tempo de residência de colônias de uma espécie de formiga predadora no Cerrado. Nós amostramos artrópodes epigeicos usando armadilhas de queda e medimos a profundidade do folhiço na borda (<15 m da borda do fragmento) e interior (>45 m da borda do fragmento) de dois fragmentos de Cerrado (fisionomia de cerradão; formação florestal com árvores de 10 a 12 m de altura). As amostragens foram feitas duas vezes: uma na estação chuvosa e outra na seca. Nós também avaliamos o tempo de residência de colônias de Odontomachus chelifer (Formicidae: Ponerinae) depois de um ano nos mesmos locais de estudo. Nós comparamos: composição e diversidade beta das comunidades de artrópodes (no nível de ordem), formigas (no nível de espécie) e grupos funcionais de formigas; riqueza e diversidade de espécies de formigas; profundidade do folhiço; e tempo de residência de colônias de O. chelifer. Nenhuma mudança nas comunidades de artrópodes, formigas ou grupos funcionais foi detectada nas bordas em comparação ao interior de fragmentos em nenhuma estação ou fragmento amostrados. A profundidade do folhiço também não sofreu efeito de borda, assim como o tempo de residência de colônias de O. chelifer. Nosso estudo contribui no conhecimento de efeitos de borda mostrando que a fauna de artrópodes em alguns fragmentos de Cerrado não sofre efeitos de borda (pelo menos que sejam detectáveis na escala de um ano). No Cerrado, efeitos de borda parecem ocorrer segundo uma dinâmica diferente em comparação a outros biomas de maior complexidade estrutural da vegetação. Além disso, efeitos de borda no Cerrado parecem sujeitos a variação temporal, que é raramente levada em conta na maior parte dos estudos sobre efeitos de borda. Estudos adicionais devem investigar os mecanismos geradores e a dinâmica temporal de efeitos de borda no Cerrado, de maneira a clarear o quão pervasivos eles são neste bioma.
Palavras-chave: efeitos de borda, fragmentação de habitat, artrópodes, formigas, grupos funcionais, ecologia de comunidades.
consequences are edge effects. Edge effects occur within abrupt transition areas between adjunct ecosystems and may have several negative effects on ecological processes. The increase in land use for human activity has driven the Brazillian Cerrado savanna to an increasingly fragmented landscape, subjecting the remaining areas to edge effects, with consequent disruption of ecosystem processes performed by arthropods. This study assessed edge effects on epigaeic arthropods, leaf litter, and colony residence time of a predatory epigaeic ant species in the Cerrado. We sampled epigaeic arthropods using pitfall traps and measured leaf litter depth at the edge (<15 m from the fragment edge) and in the interior (>45 m from the fragment edge) of two Cerrado fragments (“cerradão” physiognomy; forest with more or less merging canopy, 10-12 m tall). Samplings were carried out once in the rainy and once in the dry season. At the edge and interior of each site, we also assessed colony residence time of Odontomachus chelifer (Formicidae: Ponerinae) after one year. We compared community composition and beta diversity of arthropods (at the order level), ants (at the species level), and ant functional groups; ant species richness and diversity; leaf litter depth; and O. chelifer colony residence time. No change in arthropod, ant or functional communities was detected between edges and interior of any fragment or season sampled. Leaf litter depth did not suffer edge effects as well; neither did O. chelifer colony residence time. Our study contributes to the knowledge on edge effects by showing that the arthropod fauna in some Cerrado fragments are not affected at edges (at least at 1-year scale). In Cerrado, edge effects seem to occur according to a different dynamic in comparison to other biomes with greater structural complexity of the vegetation. Besides, edge effects in the Cerrado seem to be subject to temporal variation, which is rarely taken into account in most studies on edge effects. Further studies should investigate causal mechanisms and temporal dynamics of edge effects in the Cerrado in order to evaluate how pervasive they are in this biome.
Keywords: edge effects, habitat fragmentation, arthropods, ants, functional groups, community ecology.
Habitat fragmentation and edge effects ... 5
Arthropods and anthropogenic disturbances ... 6
The fragmentation of the Cerrado ... 7
Objectives ... 8
MATERIALS AND METHODS ... 9
Study areas ... 9
Odontomachus chelifer ... 11
Sampling design ... 12
Ant functional groups ... 13
Data analyses ... 13
Community composition ... 13
Ant diversity ... 14
Litter depth ... 14
Colony residence time of O. chelifer ... 15
RESULTS ... 15
Community composition ... 16
Ant diversity ... 18
Litter depth ... 21
Colony residence time of O. chelifer ... 22
DISCUSSION ... 23 REFERENCES ... 27 APPENDICES... 36 Appendix 1 ... 36 Appendix 2 ... 37 ANEXES ... 42
INTRODUCTION
Habitat fragmentation and edge effects
Habitat fragmentation is currently one of the greatest threats to biodiversity worldwide (Saunders et al. 1991, Wilson et al. 2016). Some regions of the Amazon forest, for example, lost an average of ca. 13000 km² of forest area per year between 2000 and 2002, with a great many portions of the remaining areas comprised of forest fragments (Broadbent et al. 2008). By reducing habitat amount, isolating patches of remaining vegetation and increasing edge area in remaining fragments (Haddad et al. 2015), habitat fragmentation may have several effects on biodiversity, including changes to community composition, population distribution, genetic diversity, trophic chain lengths, species interactions, breeding success, dispersal success, predation rates and animal behavior (Fahrig 2003, Ewers and Didham 2006, Laurance et al. 2011). One of the main drivers of these changes are edge effects.
Edge effects occur within abrupt transition areas between adjunct ecosystems (Murcia 1995). In the case of human-modified landscapes, forest fragments are usually left surrounded by matrices of low vegetation cover, such as plantations and pastures (Melo et al. 2013). This usually leaves forest edges subject to influxes of energy: light, heat and wind penetrate the edges and change local abiotic conditions (Ries et al. 2004), which often end up with higher temperatures and lower humidity levels than the interior of fragments (Matlack 1993, Didham and Lawton 1999, Laurance et al. 2011, Christianini and Oliveira 2013). These altered abiotic conditions may give rise to a myriad of consequences for organisms in habitat edges (Murcia 1995, Ries et al. 2004, Ewers and Didham 2006, Haddad et al. 2015).
Edge effects can have cascading effects throughout ecosystems. In tropical rainforests, altered abiotic conditions at fragment edges change vegetation structure (Didham and Lawton 1999, Nascimento and Laurance 2004) and cause shifts in floristic composition due to higher recruitment of pioneer plant species and elevated mortality of old-growth tree species (Laurance et al. 2006, Tabarelli et al. 2008, but see Williams-Linera 1990). Ultimately, ecosystem processes to which plants are determinant get disrupted. Leaf litter structure, for example, may change in various ways at habitat edges (Didham 1998; Didham and Lawton 1999; Delgado et al. 2013). Edge effects can then cascade even further through changes in leaf litter characteristics, which in turn can markedly affect arthropod communities (Bultman and Uetz 1984; Oliver et al. 2000; Silva et al. 2011, Delgado et al. 2013).
Animals can be affected by edge effects in several ways. For instance, changes in distribution (Bolger et al. 2000, Wirth et al. 2007, Meyer et al. 2009), activity patterns (Clopton and Gold 1993), reproductive behavior (Bellinger et al. 1989), predation rates (Paton
1994) and competition (Holway 2005) have already been documented for many animal taxa. Yet, understanding how animals respond to edge effects is difficult since individual species may respond differently to habitat edges (Ries et al. 2004, Ewers and Didham 2006): different species of beetles may show preference for forest edge or interior sites in Amazonia (Didham et al. 1998) and in Atlantic rainforest (Filgueiras et al. 2016b); butterflies with contrasting life histories may show different behavior towards edges in prairies (Ries and Debinski 2001) and tropical forests (Filgueiras et al. 2016a); some species of ants are more abundant than others in edges of Cerrado fragments (Brandão et al. 2011). The different ways in which edges affect individual animal species commonly result in changes to community composition at these locations (Didham et al. 1998, Carvalho and Vasconcelos 1999, Ries et al. 2004, Ewers and Didham 2006, Brandão et al. 2011, Delgado et al. 2013, Filgueiras et al. 2016a, Filgueiras et al. 2016b).
Arthropods and anthropogenic disturbances
Arthropods comprise the most important terrestrial animal group in terms of abundance, biomass and diversity (Wilson 1987, Stork 1988). They are also responsible for ecosystem functioning in diverse ways. Bees, for example, are the main insects responsible for the pollination of crops and wild plants (Potts et al. 2010). Termites are responsible for several ecosystem processes in the soil, such as regulating the distribution of organic matter and plant nutrients and influencing soil water regimes (Stork and Eggleton 1992). Several other taxa play major roles in other ecosystem functions, participating massively in nutrient cycling, trophic webs, seed dispersal, ecosystem engineering and a plethora of interactions with both other animals and plants (Stork and Eggleton 1992, Miller 1993, Prather et al. 2013). These features, combined with their relative ease of sampling and identification, make arthropods ideal organisms for monitoring ecological disturbances (Kremen et al. 1993). Indeed, several studies show that they respond to habitat fragmentation and can be efficiently employed in environmental assessment (Kremen et al. 1993, McGeoch 1998, Gibb and Hochuli 2002, Andersen and Majer 2004, Uehara-Prado et al. 2009), even at the order level (Bolger et al. 2000, Haskell 2000).
Within arthropods, one group stands out due to its numeric dominance and widespread occurrence throughout several ecosystems: the ants. Alongside termites, ants comprise one third of the animal biomass in the Amazon forest (Fittkau and Klinge 1973). They are typically regarded as ecosystem engineers due to their great influence on physical and chemical aspects of the soil, as they commonly nest in it (Folgarait 1998). However, they
are also important in several ecological interactions. Leaf-cutting ants, for example, exert considerable herbivory rates on tropical forests, and are able to harvest tons of leaf biomass in one year (Herz et al. 2007). Ants are also important seed dispersers for many plant species (Rico-Gray and Oliveira 2007, Christianini and Oliveira 2009). They also maintain diverse trophic interactions with other animals, such as mutualisms with honeydew-producing insects (Rico-Gray and Oliveira 2007) and predation on many arthropod taxa (Schultz and McGlynn 2000). As with other invertebrates, ants are regarded as useful organisms to assess ecological disturbance due to their ubiquity, great functional diversity, ease of sampling and sensitiveness to environmental changes (Andersen and Majer 2004, Underwood and Fisher 2006). Many studies show that ant communities show shifts in their composition in fragmented landscapes, due to disturbance-adapted species being replaced by vulnerable ones in disturbed areas (Carvalho and Vasconcelos 1999, Hoffmann and Andersen 2003, Underwood and Fisher 2006, Crist 2009, Leal et al. 2012).
One classification of biodiversity that has grown considerably in recent years is the adoption of functional groups (Petchey and Gaston 2006). Classifying species into groups in regard to their morphology (Moretti et al. 2017), natural history (Hawkins and MacMahon 1989) or ecosystem function (Bengtsson 1998) allows linking measures of diversity to ecosystem functioning (Díaz and Cabido 2001). It also holds potential as a tool for assessing human impacts on ecosystems (Chapin et al. 2000) or communities (Andersen and Majer 2004). For example, the most used functional group classification of ants assigns ant genera into groups based on their responses to environmental stress and disturbance, and it has shown considerable predictive capacity as to the responses of each group to anthropogenic impacts (Andersen 1995): generalist groups commonly thrive in disturbed areas and outcompete other, more specialized groups (Gómez et al. 2003, Hoffmann and Andersen 2003, Andersen and Majer 2004, Bestelmeyer and Wiens 1996, Crist 2009, Leal et al. 2012, Pacheco et al. 2017).
The fragmentation of the Cerrado
The Cerrado is a neotropical savanna-like vegetation which covers ca. 2 million km², 22% of the land area of Brazil (Ratter et al. 1997). It comprises physiognomies ranging from open fields to dense forests and harbors an estimated number of 160000 species of plants, animals and fungi (Oliveira and Marquis 2002). This savanna suffers great anthropogenic interference: between the years 1990 and 2010, land use for production of soybean (one of the main crops cultivated in Brazil) increased from 4.6 to 12.4 million ha
(Dias et al. 2016). Habitat conversion for agriculture and pasture is one of the main threats to the Cerrado, and have driven it towards increasingly fragmented landscapes (Carvalho et al. 2009, Dias et al. 2016), subjecting the remaining areas to edge effects (Melo et al. 2013).
Edge effects have been more commonly studied in tropical rainforests (Laurance 2004, Tabarelli et al. 2008, Laurance et al. 2011), which are expected to be more susceptible to the energy flows (which are the main drivers of edge effects) than biomes with lower structural complexity of the vegetation (Murphy and Lugo 1986, Benítez-Malvido 2014, Mendonça et al. 2015, Moreno et al. 2014, Arruda and Eisenlohr 2016). Some studies in dry forests, for example, did not find edge effects concerning microclimatic variables (Mendonça et al. 2015, Arruda and Eisenlohr 2016), litter decomposition (Moreno et al. 2014), canopy cover (Arruda and Eisenlohr 2016), and small-mammal assemblages (Napoli and Caceres 2012). However, edge effects associated with abiotic factors (Dodonov et al. 2013), plants (Christianini and Oliveira 2013, Dodonov et al. 2013, Mendonça et al. 2015), beetles (Martello et al. 2016), ants (Brandão et al. 2011, Christianini and Oliveira 2013), ant-plant interactions (Christianini and Oliveira 2013), and litter biomass (Dodonov et al. 2016) have already been reported in Cerrado.
This study investigates whether epigaeic arthropods are subject to edge effects in the Cerrado. Edge effects involving insects have been poorly studied in Cerrado (Brandão et al. 2011, Christianini and Oliveira 2013, Martello et al. 2016). Moreover, since the responses of organisms and ecosystem processes to edge effects are markedly variable (Ries et al. 2004), there is pressing need for more studies on this subject. Because habitat fragmentation disrupts the ecosystem processes in which arthropods participate (Didham et al. 1996), a proper understanding on how arthropods (and especially ants) can be affected in increasingly disturbed ecosystems is crucial for the maintenance of viable ecological communities (Crist 2009, Christianini et al. 2014).
Objectives
We investigated the following questions: 1) are epigaeic arthropod communities subject to edge effects in the Cerrado?; 2) are epigaeic ant communities subject to edge effects in the Cerrado? 3) does leaf litter depth change in Cerrado fragment edges?; 4) does colony residence time in the ant Odontomachus chelifer differ between edge and interior of Cerrado fragments?
In order to answer these questions, we sampled epigaeic arthropod communities at the edge and in the interior of Cerrado fragments to evaluate possible changes to their
composition resulting from edge effects. We also recorded leaf litter depth in both locations, since changes in this layer can affect ground-dwelling arthropod communities (Haskell 2000, Delgado et al. 2013) and other ecosystem processes (Didham 1998 and references therein). Colony residence time of the predatory ant Odontomachus chelifer (Formicidae: Ponerinae) was followed through one year at the edge and interior of these fragments. A previous work (Christianini and Oliveira 2013) has already provided evidence that colonies of this ant survive less at the edge than in the interior of one Cerrado fragment.
We sampled two fragments that differ in floristic composition and that are surrounded by different matrix types. We expected that edge effects would differ between these fragments, since other studies have already shown that fragmentation effects may differ between fragments with different vegetation or matrix types (e.g. Ricketts 2001, Ewers and Didham 2006, Delgado et al. 2013, Dodonov et al. 2013). We also conducted samplings in different seasons because we expected that climate variation typical of Cerrado (Oliveira-Filho and Ratter 2002) might lead to different edge effects on the arthropod communities (e.g. Barbosa and Marquet 2002).
We analyzed the arthropod communities at the order level and the ant community at the species and functional group levels. To our knowledge, this is the first assessment of edge effects in Cerrado involving ant functional groups and several arthropod orders.
MATERIALS AND METHODS Study areas
Given that the Cerrado is a highly variable biome with diverse physiognomies, we chose to focus on the cerradão, which is a physiognomy consisting of 50-90% canopy cover and trees of 8-12 m tall (Oliveira-Filho and Ratter 2002). Both Cerrado fragments are located in the state of São Paulo, southeast Brazil. Local climate is characterized by two well-defined seasons: a cold and dry season from April to September (hereafter “dry season”) and a hot and wet season from October to March (hereafter “rainy season”).
One of the fragments (hereafter “Itirapina”) is located in the natural reserve Estação Experimental de Itirapina (22°12’S, 47°51’W) (Fig. 1). Local mean annual temperature is 21.9°C, with maximum and minimum temperatures reaching 24.9°C (January) and 17.8°C (June), respectively. Mean annual rainfall is 1459 mm, with highest and lowest precipitation rates in February (275 mm) and July (24 mm), respectively. The study area is a 177 ha fragment of Cerrado cut by fire breaks dominated by herbs and grasses, and surrounded by pastures and Pinus sp. and Eucalyptus sp. plantations. Main plant species in
the site include Miconia rubiginosa (Bonpl.) DC., Xylopia aromatica (Lam.) Mart., Pouteria
torta (Mart.) Radlk, Syagrus petraea (Mart.) Becc., Attalea geraensis Barb. Rodr., and Bromelia balansae Mez (Giannotti 1998). Cerrado edges sampled in this fragment face the
fire breaks.
The other fragment (hereafter “Mogi-Guaçu”) is located in the Biological Reserve of Mogi-Guaçu (22° 18’S, 47° 11’W) (Fig. 1). Local mean annual temperature is 20.6°C, with maximum and minimum temperatures ranging from 23.5°C (February) to 16.3°C (July), respectively, and mean annual rainfall is 1352 mm, with highest and lowest precipitation rates in January (235.5 mm) and August (30.4 mm), respectively. The study area is a 343.42 ha fragment surrounded by roads, farms and Pinus sp. and Eucalyptus sp. plantations. Local vegetation is dominated by Siparuna guianensis (Aubl.) Tulasne, but representatives of
Miconia albicans Triana, Qualea grandiflora Mart., Anadenanthera falcata (Benth.) Speg.
and other common Cerrado plant species are also present (Mantovani and Martins 1993). Edges sampled in this site face an unpaved road.
Fig. 1: Itirapina (A) and Mogi-Guaçu (B) fragments (yellow lines) and their location within the state of São Paulo (inset map). Source: Esri; DigitalGlobe, GeoEye, Earthstar Geographics, CNES/Airbus DS, USDA, USGS, AeroGRID, IGN, and the GIS User Community.
Odontomachus chelifer
Odontomachus chelifer is an epigaeic ant species widely distributed in the
Neotropics (Brown 2000). Although it feeds mostly on arthropods (Fowler 1980, Raimundo et al. 2009), this ant is an important seed disperser of many plant species in the Cerrado
(Christianini et al. 2012 and references therein). Here, we followed the residence time of O.
chelifer colonies both in the Itirapina and in the Mogi-Guaçu fragments, so as to understand if
this species’ response to fragment edge depends on the study site.
Sampling design
Arthropods were sampled with pitfall traps consisting of plastic cups (5 cm in diameter x 6 cm in depth) filled with 70% ethanol, detergent drops, and salt. Samplings were made in blocks consisting of two parallel transects, one in the fragment edge (<15 m from the fragment edge) and another in the fragment interior (>45 m from the fragment edge; following criteria by Christianini and Oliveira 2013). In each transect we set eight pitfalls 37 ± 10 m (mean ± SD) apart from one another, left open for 24h. Two blocks were established in Itirapina and three in Mogi-Guaçu. Blocks were 1300 ± 282 m (mean ± SD) apart from one another. Samplings took place in the rainy (January) and dry seasons (July) of 2016. A total of N = 160 traps were set (80 pitfalls per season: Mogi-Guaçu: 8 pitfalls per transect x 2 transects (edge/interior) x 3 blocks = 48; Itirapina = 8 pitfalls per transect x 2 transects (edge/interior) x 2 blocks = 32). Trapped arthropods were taken to the laboratory for sorting and identification. Arthropods were identified to the order level, and ants to the species or morphospecies level.
Litter depth measurements were taken with a graded metal tube at eight sampling points in the same transects used for pitfall sampling. At each point we took three measures (1 m apart from one another) in four directions perpendicular to one another, totaling 12 measurements for each point. The mean of these 12 measurements was used in the analyses. These measurements were also made twice, once in the rainy (January) and once in the dry season (July) of 2016 (e. g. Sizer et al. 2000). Thus, out of a total of 1920 measurements taken, N = 160 points were used in the analyses (Mogi-Guaçu: 8 points per transect x 2 transects (edge/interior) x 3 blocks = 48; Itirapina: 8 points per transect x 2 transects (edge/interior) x 2 blocks = 32).
During the rainy season, we tagged O. chelifer colonies in the same blocks used for pitfall sampling. After 1 year, colonies were checked for activity, which we used as an indicator of colony residence time. The nest entrance was poked with a stick, and if no ant activity was detected in the nest or within 30-cm around it, the colony was considered inactive. In each block we tagged 6 nests, thus making a total of N = 60 nests (Mogi-Guaçu: 6 nests per transect x 2 transects (edge/interior) x 3 blocks = 36; Itirapina: 6 nests per transect x 2 transects (edge/interior) x 2 blocks = 24).
Ant functional groups
We used a classification of ant functional groups based on the work of Leal and collaborators (2012), which successfully employed these functional groups in analyzing effects of forest fragmentation on ant assemblages in the Atlantic rainforest. This classification assigns ant genera to the following functional groups: 1) Cryptic Predators – small and minute species specialized on preying arthropods; 2) Cryptic Omnivores – small and minute generalist species; 3) Epigaeic Predators – medium and large predators specialized in the predation of arthropods; 4) Epigaeic Omnivores – medium and large generalist or scavenger species; 5) Arboreal Dominants – highly aggressive species which nest in trees; 6) Arboreal Subordinates – other ant species which also nest in trees; 7) Opportunists – generalist and poorly competitive species; 8) Army Ants – nomadic species which recruit in huge numbers; 9) Leaf-Cutting Attini – Atta and Acromyrmex genera, highly specialized and polymorphic ants which cultivate fungus using cut leaves; 10) Non Leaf-Cutting Attini – monomorphic ant species with small colonies which also cultivate fungus, but do not cut leaves for doing so.
Data analyses
We hereafter exclude ants and termites from what we call “arthropod community”, and refer to the ant community classified at the species level using the terms “ant community” or “ant species”, and to the ant community classified at the functional group level using “functional community” or “functional groups”. Ants and termites were excluded from the arthropod community because diversity analyses using total or mean abundances should not be performed with social organisms (Gotelli et al. 2011).
Community composition
We used permutation tests to compare beta diversity between fragment edge and interior. In this context, beta diversity is considered as the variability in community composition in a given area, and it is calculated as the average distance of group members to the group centroid in a multivariate space (Anderson et al. 2006). We used this diversity measure since we expected that different arthropod orders/ant species/ant functional groups would respond in different ways to habitat edges. This could cause communities in fragment edges to have more variable composition than communities in the fragment interior (e.g. Didham et al. 1998). This effect would not be detected in multivariate analyses which test for
differences in centroid location, such as permANOVA (see below). Calculating beta diversity in this way thus serves two purposes: 1) assessing beta diversity of communities; 2) checking multivariate homogeneity of variances before running a permANOVA (Anderson et al. 2006). We constructed NMDS graphs with 1000 restarts for visual inspection of the arthropod, ant and functional communities. The arthropod community was analyzed by pooling the mean abundance from all traps within blocks, while ant and functional communities were analyzed with relative abundances, that is, the count of incidences of each species in samples within a block. We then performed permANOVAs with 999 permutations in order to test for edge effects on the sampled communities. Permutations were ran only between study fragments, in order to incorporate between-fragment variability in the analysis. Rainy and dry seasons were treated separately for all these three analyses, and all of them were performed with the Morisita-Horn dissimilarity index, which gives more value to abundant species, and as such is: 1) resistant to under-sampling; 2) a good index to understand functional differences between ecosystems, given that ecological processes are often more influenced by the most abundant species (Jost et al. 2011, Chao et al. 2014).
Ant diversity
We used Hill numbers to compare ant species richness and diversity. Hill numbers integrate species richness and abundance into a class of measures differing only by an exponent q. These measures have several statistical advantages over other diversity indexes (Chao et al. 2014). We constructed rarefaction curves using species richness (q = 0), and calculated the inverse of Simpson concentration (q = 2) (Chao et al. 2014). We chose the Simpson diversity measure as it is, just as the Morisita-Horn index, more influenced by the most abundant species (Jost et al. 2011). Diversity indices were calculated with 1000 boostrap replications of relative abundances. Separate curves/indexes were calculated for each fragment and season sampled. All curves were extrapolated to twice the number of sampling units in each fragment, so as to detect if: 1) our sampling effort was sufficient to detect most ant species in our fragments; 2) additional sampling would have yielded different curves from those obtained with the employed sampling.
Litter depth
We tested for differences in litter depth between fragment edge and interior using a GLMM with a quasi-Poisson family and log link function. Fixed effects were edge effect, fragment (Itirapina/Mogi-Guaçu) and the interaction between edge effect and fragment.
Blocks were incorporated as a random effect. Parameter estimates were performed with maximum likelihood. We then used Spearman’s rank correlation coefficient (ρ) to test for correlations between: 1) litter depth and total arthropod abundance; 2) litter depth and ant richness. We did so since litter acts as a microclimatic refuge to epigaeic invertebrates (Bultman and Uetz 1984 and references therein), and thus it could offset edge effects derived from harsh abiotic conditions, hiding differences in arthropod or ant communities at fragment edges. Thus, investigating a correlation between litter depth and arthropod abundance or ant richness would suggest an explanation for differences in arthropod/ant communities at fragment edges (Haskell 2000, Delgado et al. 2013). We chose Spearman’s ρ as it does not assume: 1) a linear relationship between variables; 2) bivariate normality (Quinn and Keough 2002). Both these analyses were conducted separately for each season and fragment sampled.
Colony residence time of O. chelifer
Colony residence time of O. chelifer was analyzed using a GLMM with Binomial family and logit link function. The response variable, colony activity (active/inactive), was modeled in response to the predictor variables edge effect, fragment and the interaction between edge effect and fragment. Blocks were incorporated as a random effect. Parameter estimates were performed with Laplace approximation and maximum likelihood. The dispersion parameter defined in the model was checked and considered acceptable for the estimation of p values.
All analyses were performed using the R language version 3.2.3 (R Core Team, 2015). Rarefaction curves were made with the iNEXT (Hsieh et al 2016) and ggplot2 packages (Wickham, 2009), NMDS, permANOVA and beta diversity (betadisper function) were performed with the vegan package (Oksanen et al 2016), the quasi-Poisson GLMM was ran with the MASS package (Venables and Ripley 2002) and the Binomial GLMM was ran with the lme4 package (Bates et al 2015).
RESULTS
We sampled a total of 1206 and 1186 arthropods from 10 and 11 orders in the rainy and dry seasons (Appendix 1). Ant abundance was represented by 3019 and 2409 individuals in the rainy and dry seasons, respectively, which corresponded to 59 and 45 ant species and 9 and 7 functional groups (Appendix 2).
Community composition
No differences in beta diversity between edge and interior sites were found for any community or season sampled (Table 1).
Table 1: Beta diversity (measured as average distance of group members to group centroid in multivariate space) of arthropod, ant and functional communities at fragment edge and interior of cerrado in the rainy and dry seasons.
Community Season Edge/Interior Beta diversity Pairwise permuted p-value
Arthropods Rainy Edge 0.068 0.494
Interior 0.088 -
Arthropods Dry Edge 0.138 0.235
Interior 0.078 -
Ants Rainy Edge 0.218 0.558
Interior 0.251 -
Ants Dry Edge 0.176 0.610
Interior 0.195 -
Functional groups Rainy Edge 0.042 0.550
Interior 0.036 -
Functional groups Dry Edge 0.053 0.240
Interior 0.073 -
NMDS analyses did not show a clear separation between cerrado fragment edge and interior for any community or season (Fig. 2). Indeed, permANOVAs did not detect an edge effect on the sampled communities in any season (Table 2).
Fig. 2: NMDS analyses of the sampled communities in cerrado. Circles: edge blocks; triangles: interior blocks. Stress values: Arthropod community, rainy season = 0.05; Arthropod community, dry season = 0.09; Ant community, rainy season = 0.04; Ant community, dry season = 0.07; Functional community, rainy season = 0.07; Functional community, dry season = 0.09.
Table 2: permANOVA results of edge effects on the arthropod, ant and functional communities in cerrado, for the rainy and dry seasons.
Community Season Predictor variable
DF F R² Pr(>F)
Arthropods Rainy Edge effect 1 -0.076 -0.009 0.884
Residuals 8 - 1.009 -
Total 9 - 1 -
Arthropods Dry Edge effect 1 0.447 0.052 0.224
Residuals 8 - 0.947 -
Total 9 - 1 -
Ants Rainy Edge effect 1 0.190 0.023 0.633
Residuals 8 - 0.976 -
Total 9 - 1 -
Ants Dry Edge effect 1 0.858 0.096 0.160
Residuals 8 - 0.903 -
Total 9 - 1 -
Functional groups Rainy Edge effect 1 0.673 0.077 0.377
Residuals 8 - 0.922 -
Total 9 - 1 -
Functional groups Dry Edge effect 1 2.542 0.241 0.130
Residuals 8 - 0.758 -
Total 9 - 1 -
Ant diversity
Rarefaction curves of the ant community are shown in Figure 3. Extrapolation of the curves show that more ant species would have been sampled with additional sampling effort, yet ant species richness would not reach an asymptote for any fragment, nor would our results have changed, even with twice the sample size we employed. Overlap of 95% confidence intervals indicates that ant species richness did not differ significantly between fragment edge and interior of cerrado, for any season or fragment (Fig. 3). Ant diversity measures are shown in Table 3. Neither ant richness nor Simpson diversity differed between edge and interior sites for any cerrado fragment and seasons sampled, as indicated by the overlap of the 95% confidence intervals.
Fig. 3 Rarefaction curves constructed for the ant community sampled in cerrado fragments in São Paulo. Vertical axis shows species richness calculated using Hill numbers (q = 0). Triangles/circles indicate the point where the sample size of each fragment edge/interior was reached (N = 24 in Mogi-Guaçu and N = 16 in Itirapina). Dashed lines are the extrapolation of curves up to twice the number of sampling units of each fragment edge/interior. Overlap of confidence intervals (darker shaded areas) indicates that species richness in fragment edge and interior of cerrado fragments does not differ significantly. Curves marked with a triangle: fragment interior; curves marked with a circle: fragment edge.
Table 3: Ant community diversity estimates using Hill numbers (q = 0, 2). Overlap of all confidence intervals demonstrates that diversity measures did not differ between edge and interior of cerrado fragments.
Fragment Season Edge / Interior Diversity measures Observed Estimator Estimated SE 95 % CI
Mogi-Guaçu Rainy Edge Species richness 32 52.245 15.774 37.248 - 110.092
Simpson 16.294 17.686 1.519 16.294 - 20.662
Interior Species richness 39 64.875 16.919 47.036 - 122.314
Simpson 20.451 23.058 1.98 20.451 - 26.938
Dry Edge Species richness 32 51.805 15.445 37.129 - 108.469
Simpson 19.114 21.675 1.913 19.114 - 25.424
Interior Species richness 31 38.5 5.701 32.996 - 59.175
Simpson 16.889 19.385 2.17 16.889 - 23.638
Itirapina Rainy Edge Species richness 28 39.596 9.348 30.897 - 74.41
Simpson 12.153 13.102 1.39 12.153 - 15.827
Interior Species richness 25 77.708 33.515 41.822 - 190.15
Simpson 12.505 13.639 1.339 12.505 - 16.263
Dry Edge Species richness 24 33.492 8.489 26.113 - 66.648
Simpson 14.327 16.71 1.887 14.327 - 20.408
Interior Species richness 26 31.424 4.765 27.232 - 49.881
Litter depth
GLMM results for litter depth are shown in Table 4. Although there was a significant effect of cerrado fragments in the dry season, no significant edge effect or interaction between edge and fragment was detected in litter depth. There was no correlation between total arthropod abundance and litter depth between edge and interior of cerrado fragments for any fragment or season (Table 5). Ant species richness also did not correlate with litter depth at any fragment edge or season, although a marginally nonsignificant correlation occurred at the edge of the Itirapina fragment during the rainy season (Table 5).
Table 4: Results of the quasi-Poisson GLMM for litter depth for both seasons sampled. The asterisk (*) indicates an interaction between predictor variables.
Rainy Season Random factor SD
Transect (block) 1.105
Response variable Value SE t Pr(>|t|)
Intercept 0.971 0.252 3.840 <0.001
Edge effect 0.110 0.219 0.502 0.616
Fragment 0.179 0.208 0.858 0.393
Edge effect*Fragment -0.124 0.284 -0.438 0.662
Dry Season Random factor SD
Transect (block) 0.714
Response variable Value SE t Pr(>|t|)
Intercept 1.492 0.141 10.536 <0.001
Edge effect -0.104 0.121 -0.857 0.393
Fragment -0.239 0.119 -2.002 0.048
Table 5: Spearman’s rank correlation coefficient between: 1) litter depth and arthropod abundance; 2) litter depth and ant species richness between edge and interior for both cerrado fragments and seasons sampled.
Correlated variable Fragment Season Edge / Interior
Spearman’s ρ p-value
Arthropod abundance Mogi-Guaçu Rainy Edge 0.140 0.511 Interior <0.001 0.997
Dry Edge 0.272 0.197
Interior 0.005 0.981
Itirapina Rainy Edge 0.174 0.517
Interior -0.160 0.552
Dry Edge 0.245 0.358
Interior -0.068 0.800
Ant richness Mogi-Guaçu Rainy Edge 0.179 0.400
Interior -0.037 0.860
Dry Edge -0.111 0.603
Interior -0.276 0.190
Itirapina Rainy Edge -0.495 0.051
Interior -0.347 0.187
Dry Edge -0.133 0.621
Interior -0.152 0.571
Colony residence time of O. chelifer
Of the 36 O. chelifer nests tagged in the cerrado of Mogi-Guaçu, nine were inactive after one year: three (8%) in the edge and six (16%) in the interior. In Itirapina, out of 24 tagged nests, 10 were inactive: four (16%) in the edge and six (25%) in the interior (Fig. 4). GLMM results showed that neither edge nor fragment or interaction between edge and fragment influenced the residence time of nests (Table 6).
Fig. 4: Colony residence time of O. chelifer in the edge and interior of the cerrado in Mogi-Guaçu and Itirapina.
Table 6: Results of the Binomial GLMM for O. chelifer colony residence time in cerrado. The asterisk (*) indicates an interaction between predictor variables.
Random factor Variance
Transect (block) 0.956
Response variable Value SE z Pr(>|z|)
Intercept 0.713 0.749 0.939 0.425
Edge effect -0.698 0.875 -0.857 0.391
Fragment 0.906 1.340 0.952 0.341
Edge effect*Fragment -0.223 1.253 -0.302 0.763
DISCUSSION
Edge effects have so far been neglected in the Cerrado: a search in Web of Science using the terms “edge effects” and “Cerrado” returns only 20 results. Our study adds to the still incipient knowledge on edge effects in this increasingly disturbed savanna. We demonstrate that epigaeic arthropod and ant communities, litter depth and O. chelifer colony residence time did not differ between the edge and interior of the Cerrado fragments we sampled. Current knowledge on edge effects is heavily based upon studies in tropical rainforests (Laurance 2004, Laurance et al 2011), and we believe their dynamics may be different in forests with less complex vegetation, such as dry forests (Murphy and Lugo 1986,
Benítez-Malvido et al. 2014, Mendonça et al. 2015, Moreno et al. 2015, Arruda and Eisenlohr 2016).
Responses of arthropods to habitat edges are highly variable, because species with different life-histories are affected differentially by habitat edges (Didham et al. 1998; Ewers and Didham 2006; Filgueiras et al. 2016; Ries and Debinski 2001; Holway 2005). This commonly translates into community composition in fragment edges being different from that in the interior (Didham et al. 1998, Carvalho and Vasconcelos 1999, Ries et al. 2004, Ewers and Didham 2006, Brandão et al. 2011, Delgado et al. 2013, Filgueiras et al. 2016a, Filgueiras et al. 2016b). This effect is commonly observed in ant communities (Suarez et al. 1998, Carvalho and Vasconcelos 1999, Hoffmann and Andersen 2003, Underwood and Fisher 2006, Crist 2009, Leal et al. 2012). Anthropogenic impacts often simplify vegetation structure (Didham and Lawton 1999, Laurance et al. 2011), bringing about harsh abiotic factors (Sizer and Tanner 1999) and lower resource availability (McGlynn 2006), which have negative effects on ants with narrow thermal tolerances and specialized feeding, while having positive effects on more generalist species (Andersen and Majer 2004, Crist 2009, Leal et al. 2012). Specialist taxa are therefore replaced by generalist ones, and so disturbances ultimately lead to diversity reduction, a process known as biotic homogenization (Solar et al. 2015). Indeed, ant diversity has already been shown to decrease with increasing levels of disturbance (Roth et al. 1994, Abensperg-Traun et al. 1996), and studies in Amazonia (Carvalho and Vasconcelos 1999), Atlantic Forest (Leal et al. 2012), and Cerrado (Brandão et al. 2011) show that disturbed sites differ in ant species composition from undisturbed sites (see also Underwood and Fisher 2006, Crist 2009). This pattern is also observed in studies using ant functional groups. For instance, Andersen (1995) classified ant genera into functional groups that respond consistently to disturbance and stress in several studies. Generalist groups such as “Opportunists” and “Dominant Dolichoderinae” usually respond positively to disturbances, whereas specialized groups “Cryptic Species” and “Cold Climate Specialists” commonly show strong negative responses to disturbances (Hoffmann and Andersen 2003).
There was no change in litter depth at the fragment edges we sampled either, and litter depth did not explain a lack of edge effects on arthropod abundance or ant richness. Litter depth, just as invertebrate communities, changes in many different ways at forest edges. In the Amazon forest, for instance, litter depth has been shown to: 1) decrease both in linear and nonlinear ways with distance from the edge in different sites (Carvalho and Vasconcelos 1999); 2) increase in structurally closed edges but not in open ones (Didham and Lawton 1999). Litter fall rates have also been reported to be higher in edge sites in Amazonia,
although this effect disappeared 1.5 years after edge creation (Sizer et al. 2000). Differences in litter depth in forest edges have already been suggested as a cause for: 1) reduction in invertebrate abundance and ant richness near roads in a temperate forest (Haskell 2000); 2) changes in ant community composition in the Amazon forest (Carvalho and Vasconcelos 1999). In the Atlantic forest, a study has found that number of leaves in the leaf litter is positively correlated with ant richness in forest interior, but not in forest edge (Silva et al. 2011).
Edge effects have already been recorded in dry forests. In the Chaco forest, for instance, positive edge effects have been found for herbivore and parasitoid family richness (González et al. 2015) and for abundance and species richness of leaf-cutting ants of the
Acromyrmex genus (Barrera et al. 2015). On the other hand, litter decomposition did not
suffer edge effects in a study in the Chaco forest (Moreno et al. 2014), and suffered only a weak edge effect in a study in an Australian dry forest (Hastwell and Morris 2013). In Cerrado, edge effects have been recorded for beetle (Martello et al. 2016) and ant communities (Brandão et al. 2011), although the effects depended on the type of matrix (Martello et al. 2016) and fragment (Brandão et al. 2011). Another study in Cerrado did not find any changes in small-mammal assemblage richness, diversity and composition (Napoli and Caceres 2012). Highly variable edge effects on litter biomass have also been reported in the Cerrado, with increase or decrease in biomass depending on the type of litter and fragment sampled (Dodonov et al. 2016).
Despite previous records of edge effects in Cerrado (Brandão et al. 2011, Christianini and Oliveira 2013, Dodonov et al. 2013, 2016, Mendonça et al. 2015, Martello et al. 2016), our study did not detect any edge effects in the studied fragments. This may be due to the fact that savannas have high micro-climatic variations caused by discontinuities in the canopy cover (Murphy and Lugo 1986). As such, the energy flows responsible for edge effects (Ries et al. 2004) might not apply in savanna fragments as compared to more closed, forested physiognomies (Carvalho et al. 2009, Mendonça et al. 2015). Changes in abiotic variables have already been reported in fragment edges in Cerrado (Christianini and Oliveira 2013, Dodonov et al. 2013, Mendonça et al. 2015). Yet, responses of several of these variables to edges may be weak (Mendonça et al. 2015) or absent (Dodonov et al. 2013) in some fragments.
Contrary to what we expected, O. chelifer colonies did not survive less at fragment edges in comparison to the cerrado interior. This result contrasts with that of Christianini and Oliveira (2013), who followed O. chelifer colony residence time in the
Itirapina fragment during one year and found that 92% of colonies were still alive in the fragment interior, as opposed to only 33% in the edge. Apart from location within fragment edge or interior, other factors seem to have accounted for the mortality of O. chelifer colonies in our study. We believe that edge effects in Cerrado depend on at least one factor which is rarely taken into consideration in most scientific studies on edge effects: temporal variability (Murcia 1995, Barbosa and Marquet 2002). Climatic factors may play a role in ant colony survival, which may vary significantly across years (Gordon and Kulig 1998, Sanders and Gordon 2004), and be related to large-scale climatic events (Sanders and Gordon 2004) that are known to be determinant of habitat suitability for ants (Wiescher et al. 2012). These unaccounted-for climatic variations across years might explain why we found no edge effects in the cerrado of Itirapina. In fact, data from a local climatological station show that, in 2005 and 2006 – when Christianini and Oliveira (2013) conducted their study (Christianini AV, pers. comm.) – mean temperatures in the Itirapina reserve were, respectively, 23 and 12% less and rainfall was 26 and 18% less than in 2015 and 2016 (Salles, L. F. P., unpubl. data) (when this study was conducted).
Our study suggests that edge effects may not be as widespread in Cerrado as compared to other biomes. This is not to say that edge effects do not occur in Cerrado, but that they seem to behave differently than in humid forests, where edge effects dynamics are better understood (Laurance 2004, Laurance et al. 2011). Moreover, edges of Cerrado fragments have already been shown to be susceptible to, for instance, invasion by exotic grasses (Mendonça et al. 2015, see also Prasad 2009) and disruption of secondary seed dispersal by ants (Christianini and Oliveira 2013), with expected changes in the composition of the local plant community. We believe that further research on edge effects in Cerrado must take into account their mechanistic causes and temporal dynamics (Murcia 1995) to better evaluate how pervasive they are in this threatened savanna.
REFERENCES
Abensperg-Traun M, Smith GT, Arnold GW, Steven DE (1996) The Effects of Habitat Fragmentation and Livestock-Grazing on Animal Communities in Remnants of Gimlet Eucalyptus salubris Woodland in the Western Australian Wheatbelt. I. Arthropods. J Appl Ecol 33:1281–1301
Andersen AN (1995) A Classification of Australian Ant Communities, Based on Functional Groups Which Parallel Plant Life-Forms in Relation to Stress and Disturbance. J Biogeogr 22:15–29
Andersen AN, Majer JD (2004) Ants Show the way Down Under: Invertebrates as Bioindicators in Land Management. Front Ecol Environ 2:291–298
Anderson MJ, Ellingsen KE, McArdle BH (2006) Multivariate dispersion as a measure of beta diversity. Ecol Lett 9:683–693
Arruda DM, Eisenlohr P V (2016) Analyzing the edge effects in a Brazilian seasonally dry tropical forest. Brazilian J Biol 76:169–75
Barbosa O, Marquet PA (2002) Effects of forest fragmentation on the beetle assemblage at the relict forest of Fray Jorge, Chile. Oecologia 132:296–306
Barrera CA, Buffa LM, Valladares G (2015) Do leaf-cutting ants benefit from forest
fragmentation? Insights from community and species-specific responses in a fragmented dry forest. Insect Conserv Divers 8:456–463
Bates D, Maechler M, Bolker B, Walker S (2015) Fitting Linear Mixed-Effects Models Using lme4. J Statistical Softw, 67: 1-48
Bellinger RG, Ravlin FW, McManus ML (1989) Forest edge effects and their influence on the gypsy moth (Lepidoptera: Lymantriidae) egg mass distribution. Environ Entomol
18:840–843
Belshaw R, Bolton B (1993) The effect of forest disturbance on the leaf litter ant fauna in Ghana. Biodivers Conserv 2:656–666
Bengtsson J (1998) Which species? What kind of diversity? Which ecosystem function? Some problems in studies of relations between biodiversity and ecosystem function. Appl Soil Ecol 10:191–199
Bestelmeyer BT, Wiens JA (1996) The Effects of Land Use on the Structure of Ground-Foraging Ant Communities in the Argentine Chaco. Ecol Appl 6:1225–1240.
Bolger DT, Suarez A V., Crooks KR, et al (2000) Arthropods in Urban Habitat Fragments in Southern California: Area, Age, and Edge Effects. Ecol Appl 10:1230–1248
Formicidae) as indicators of edge effects. Zoologia 28:379–387
Broadbent EN, Asner GP, Keller M, et al (2008) Forest fragmentation and edge effects from deforestation and selective logging in the Brazilian Amazon. Biol Conserv 141:1745– 1757
Brown Jr WL (2000) Diversity of Ants. In: Agosti D, Majer JD, Alonso LE, Schultz TR (eds) Ants: standard methods for measuring and monitoring biodiversity. Smithsonian
Institution Press, Washington, pp 45-79
Bultman TL, Uetz GW (1984) Effect of Structure and Nutritional Quality of Litter on Abundances of Litter-dwelling Arthropods. Am Midl Nat 111:165–172
Carvalho KS, Vasconcelos HL (1999) Forest fragmentation in central Amazonia and its effects on litter-dwellings ants. Biol Conserv 91:151–157.
Carvalho FM V, De Marco P, Ferreira LG (2009) The Cerrado into-pieces: Habitat fragmentation as a function of landscape use in the savannas of central Brazil. Biol Conserv 142:1392–1403
Chao A, Gotelli NJ, Hsieh TC, et al (2014) Rarefaction and extrapolation with Hill numbers: a framework for sampling and estimation in species diversity studies. Ecol Monogr 84:45– 67
Chapin FS, Zavaleta ES, Eviner VT, et al (2000) Consequences of changing biodiversity. Nature 405:234–42
Christianini AV, Mayhé-Nunes AJ, Oliveira PS (2012) Exploitation of Fallen Diaspores by Ants: Are there Ant-Plant Partner Choices? Biotropica 44:360–367
Christianini AV, Oliveira PS (2009) The relevance of ants as seed rescuers of a primarily bird-dispersed tree in the Neotropical cerrado savanna. Oecologia 160:735–745
Christianini AV, Oliveira PS (2013) Edge effects decrease ant-derived benefits to seedlings in a neotropical savanna. Arthropod Plant Interact 7:191–199
Christianini AV, Oliveira PS, Bruna EM, Vasconcelos HL (2014) Fauna in decline: Meek shall inherit. Science 345:1129-1129
Clopton RE, Gold RE (1993) Distribution and Seasonal and Diurnal Activity Patterns of
Eutrombicula alfreddugesi (Acari: Trombiculidae) in a Forest Edge Ecosystem. J Med
Entomol 30:47–53
Crist TO (2009) Biodiversity, species interactions, and functional roles of ants (Hymenoptera: Formicidae) in fragmented landscapes: a review. Myrmecological News 12:3–13
Delgado JD, Arroyo NL, Arévalo JR, Fernández-Palacios JM (2013) Road edge effects on litter invertebrate communities of subtropical forests. J Nat Hist 47:203–236
Dias LCP, Pimenta FM, Santos AB, et al (2016) Patterns of land use, extensification, and intensification of Brazilian agriculture. Glob Chang Biol 22:2887–2903
Díaz S, Cabido M (2001) Vive la différence: plant functional diversity matters to ecosystem processes. Trends Ecol Evol 16:646–655
Didham RK (1998) Altered leaf-litter decomposition rates in tropical forest fragments. Oecologia 116:397–406
Didham RK, Ghazoul J, Stork NE, Davis AJ (1996) Insects in fragmented forests: a functional approach. Trends Ecol Evol 11:255–260
Didham RK, Hammond PM, Lawton JH, et al (1998) Beetle species responses to tropical forest fragmentation. Ecol Monogr 68:295–323
Didham RK, Lawton JH (1999) Edge Structure Determines the Magnitude of Changes in Microclimate and Vegetation Structure in Tropical Forest Fragments. Biotropica 31:17– 30
Dodonov P, Braga AL, Harper KA, Silva Matos DM (2016) Edge influence on plant litter biomass in forest and savanna in the Brazilian cerrado. Austral Ecol 1–11. doi: 10.1111/aec.12420
Dodonov P, Harper KA, Silva-Matos DM (2013) The role of edge contrast and forest structure in edge influence: vegetation and microclimate at edges in the Brazilian cerrado. Plant Ecol 214:1345–1359
Ewers RM, Didham RK (2006) Confounding factors in the detection of species responses to habitat fragmentation. Biol Rev 81:117–142
Fahrig L (2003) Effects of Habitat Fragmentation on Biodiversity. Annu Rev Ecol Evol Syst 34:487–515
Farji-Brener AG (2001) Why are leaf-cutting ants more common in early secondary forests than in old-growth tropical forests? An evaluation of the palatable forage hypothesis. Oikos 92:169–177
Filgueiras BKC, Melo DHA, Leal IR, et al (2016a) Fruit-feeding butterflies in
edge-dominated habitats: community structure, species persistence and cascade effect. J Insect Conserv 20:539–548
Filgueiras BKC, Tabarelli M, Leal IR, et al (2016b) Spatial replacement of dung beetles in edge-affected habitats: biotic homogenization or divergence in fragmented tropical forest landscapes? Divers Distrib 22:400–409
Fittkau EJ, Klinge H (1973) On Biomass and Trophic Structure of the Central Amazonian Rain Forest Ecosystem. Biotropica 5:2–14
Folgarait P (1998) Ant biodiversity and its relationship to ecosystem functioning: a review. Biodivers Conserv 7:1221–1244
Fowler HG (1980) Populations, prey capture and sharing, and foraging of the Paraguayan ponerine Odontomachus chelifer Latreille. J Nat Hist 14:79–84
Giannotti E (1998) Composição florística e estrutura fitossociológica da vegetação de cerrado e de transição entre cerrado e mata ciliar da Estação Experimental de Itirapina (SP). Master Thesis, Universidade Estadual de Campinas
Gibb H, Hochuli DF (2002) Habitat fragmentation in an urban environment: large and small fragments support different arthropod assemblages. Biol Conserv 106:91–100
Gómez C, Casellas D, Oliveras J, Bas JM (2003) Structure of ground-foraging ant
assemblages in relation to land-use change in the northwestern Mediterranean region. Biodivers Conserv 12:2135–2146
González E, Salvo A, Valladares G (2015) Arthropods on plants in a fragmented Neotropical dry forest: A functional analysis of area loss and edge effects. Insect Sci 22:129–138 Gordon DM, Kulig A (1998) The effect of neighbours on the mortality of harvester ant
colonies. J Anim Ecol 67:141–148
Gordon DM, Kulig AW (1996) Founding, Foraging, and Fighting: Colony Size and the Spatial Distribution of Harvester Ant Nests. Ecology 77:2393–2409
Gotelli NJ, Ellison AM, Dunn RR, Sanders NJ (2011) Counting Ants (Hymenoptera: Formicidae): Biodiversity Sampling and Statistical Analysis for Myrmecologists. Myrmecological News 15:13–19
Haddad NM, Brudvig LA, Clobert J, et al (2015) Habitat fragmentation and its lasting impact on Earth’s ecosystems. Sci Adv 1:e1500052
Harrison S, Bruna E (1999) Habitat fragmentation and large-scale conservation: what do we know for sure? Ecography 22:225–232
Haskell DG (2000) Effects of Forest Roads on Macroinvertebrate Soil Fauna of the Southern Appalachian Mountains. Conserv Biol 14:57–63
Hastwell GT, Morris EC (2013) Structural features of fragmented woodland communities affect leaf litter decomposition rates. Basic Appl Ecol 14:298–308.
Hawkins CP, MacMahon JA (1989) Guilds: The Multiple Meanings Of A Concept. Annu Rev Entomol 34:423–451
Herz H, Beyschlag W, Hölldobler B (2007) Herbivory Rate of Leaf-Cutting Ants in a Tropical Moist Forest in Panama at the Population and Ecosystem Scales. Biotropica 39:482–488 Hoffmann BD, Andersen AN (2003) Responses of ants to disturbances in Australia, with
particular reference to functional groups. Austral Ecol 28:444–464
Holway DA (2005) Edge effects of an invasive species across a natural ecological boundary. Biol Conserv 121:561–567
Hsieh TC, Ma KH, Chao A (2016) iNEXT: iNterpolation and EXTrapolation for species diversity. R package version 2.0.8
Jost L, Chao A, Chazdon RL (2011) Compositional similarity and (beta) diversity. In: Magurran AE, McGill BJ (eds) Biological diversity: frontiers in measurement and assesment. Oxford University Press, New York, pp. 66-84
Kaspari M, Weiser MD (1999) The size-grain hypothesis and interspecific scaling in ants. Funct Ecol 13:530–538
Kremen C, Colwell RK, Erwin TL, et al (1993) Terrestrial Arthropod Assemblages: Their Use in Conservation Planning. Conserv Biol 7:796–808
Kumar A, O’Donnell S (2009) Elevation and forest clearing effects on foraging differ between surface - and subterranean - foraging army ants (Formicidae: Ecitoninae). J Anim Ecol 78:91–97
Laurance WF (2004) Forest-climate interactions in fragmented tropical landscapes. Philos Trans R Soc B-Biological Sci 359:345–352
Laurance WF, Camargo JLC, Luizão RCC, et al (2011) The fate of Amazonian forest fragments: A 32-year investigation. Biol Conserv 144:56–67
Laurance WF, Nascimento HEM, Laurance SG, et al (2006) Rapid decay of tree-community composition in Amazonian forest fragments. Proc Natl Acad Sci U S A 103:19010– 19014
Leal IR, Filgueiras BKC, Gomes JP, et al (2012) Effects of habitat fragmentation on ant richness and functional composition in Brazilian Atlantic forest. Biodivers Conserv 21:1687–1701
Levings SC, Windsor DM (1984) Litter Moisture Content as a Determinant of Litter
Arthropod Distribution and Abundance during the Dry Season on Barro Colorado Island, Panama. Biotropica 16:125–131
Mantovani W, Martins FR (1993) Florística do Cerrado na Reserva Biológica de Mogi-Guaçu, SP. Acta Bot Brasilica 7:33–60
Martello F, Andriolli F, de Souza TB, et al (2016) Edge and land use effects on dung beetles (Coleoptera: Scarabaeidae: Scarabaeinae) in Brazilian cerrado vegetation. J Insect Conserv 20:957-970
eastern United States. Biol Conserv 66:185–194
McGeoch MA (1998) The selection , testing and application of terrestrial insects as bioindicators. Biol Rev Camb Philos Soc 73:181–201
McGlynn TP (2006) Ants on the Move: Resource Limitation of a Litter-nesting Ant Community in Costa Rica. Biotropica 38:419–427
Melo FPL, Arroyo-Rodríguez V, Fahrig L, et al (2013) On the hope for biodiversity-friendly tropical landscapes. Trends Ecol Evol 28:462–468
Mendonça AH, Russo C, Melo ACG, Durigan G (2015) Edge effects in savanna fragments: a case study in the cerrado. Plant Ecol Divers 8:493-503
Meyer ST, Leal IR, Wirth R (2009) Persisting Hyper-abundance of Leaf-cutting Ants (Atta spp.) at the Edge of an Old Atlantic Forest Fragment. Biotropica 41:711–716
Miller JC (1993) Insect natural history, multi-species interactions and biodiversity in ecosystems. Biodivers Conserv 2:233–241
Moreno ML, Bernaschini ML, Pérez-Harguindeguy N, Valladares G (2014) Area and edge effects on leaf-litter decomposition in a fragmented subtropical dry forest. Acta Oecologica 60:26–29
Moretti M, Dias ATC, de Bello F, et al (2017) Handbook of protocols for standardized measurement of terrestrial invertebrate functional traits. Funct Ecol 31:558–567 Murcia C (1995) Edge effects in fragmented forests: implications for conservation. Trends
Ecol Evol 10:58–62
Murphy PG, Lugo AE (1986) Ecology of tropical dry forest. Annu Rev Ecol Syst 17:67–88. Napoli RP, Caceres NC (2012) Absence of edge effect on small mammals in
woodland-savannah remnants in Brazil. Community Ecol 13:11–20.
Nascimento HEM, Laurance WF (2004) Biomass dynamics in Amazonian forest fragments. Ecol Appl 14:127–138
Oksanen J, Guillaume Blanchet F, Friendly M, et al (2016) Vegan: Community Ecology Package. R package version 2.4-0
Oliveira-Filho AT, Ratter JA (2002) Vegetation Physiognomies and Woody Flora of the Cerrado Biome. In: Oliveira PS, Marquis RJ (eds) The Cerrados of Brazil: Ecology and Natural History of a Neotropical Savanna. Columbia University Press, New York, p 91-120
Oliveira PS, Marquis RJ (2002) The Cerrados of Brazil: Ecology and Natural History of a Neotropical Savanna. Columbia University Press, New York