Submitted 19 August 2016Accepted 27 February 2017Published 4 April 2017

Corresponding authorCristiane Costacristianemqcostagmailcom

Academic editorNigel Andrew

Additional Information andDeclarations can be found onpage 14

DOI 107717peerj3125

Copyright2017 Costa et al

Distributed underCreative Commons CC-BY 40

OPEN ACCESS

Variegated tropical landscapes conservediverse dung beetle communitiesCristiane Costa1 Victor Hugo F Oliveira12 Rafaella Maciel3Wallace Beiroz12 Vanesca Korasaki4 and Julio Louzada12

1 Setor de EcologiaDepartamento de Biologia Universidade Federal de Lavras Lavras MG Brazil2 Lancaster Environment Centre Lancaster University Lancaster United Kingdom3Departamento de Entomologia Universidade Federal de Lavras Lavras MG Brazil4Departamento de Ciecircncias Exatas e da Terra Universidade do Estado de Minas Gerais Frutal MG BrazilThese authors contributed equally to this work

ABSTRACTBackground Conserving biodiversity in tropical landscapes is a major challenge toscientists and conservationists Current rates of deforestation fragmentation andland use intensification are producing variegated landscapes with undeterminedvalues for the conservation of biological communities and ecosystem functioningHere we investigate the importance of tropical variegated landscapes to biodiversityconservation using dung beetle as focal taxaMethods The study was carried out in 12 variegated landscapes where dung beetleswere sampled using six pitfall traps 30m apart from each other along a transect in eachstudied landscape use and cover classesmdashLUCC (forest fragment and corridor coffeeplantation and pasture) We baited each pitfall trap with 30 g of human feces and leftopen for a 48 h period We also measured three environmental variables reflectingstructural differences among the studied classes canopy cover local vegetationheterogeneity and soil sand contentResults We collected 52 species and 2695 individuals of dung beetles We observedsignificant differences in the mean species richness abundance and biomass amongclasses with forest fragments presenting the highest values forest corridors and coffeeplantations presenting intermediate values and pastures the lowest values Regardingcommunity structure we also found significant differences among classes Canopycover was the only variable explaining variation in dung beetle species richnessabundance biomass and community structure The relative importance of spatialturnover was greater than nestedness-resultant component in all studied landscapesDiscussion This study evaluated the ecological patterns of dung beetle communitiesin variegated tropical landscapes highlighting the importance of these landscapesfor conservation of tropical biodiversity However we encourage variegation for themanagement of landscapes that have already been fragmented or as a complementaryinitiative of current conservation practices (eg protection of natural habitats andestablishment of reserves)

Subjects Agricultural Science Biodiversity Conservation Biology Ecology EntomologyKeywords Biodiversity conservation Agriculture Countryside Hedgerow Landscapemodification Scarabaeinae Forest fragments Forest corridors

How to cite this article Costa et al (2017) Variegated tropical landscapes conserve diverse dung beetle communities PeerJ 5e3125 DOI107717peerj3125

INTRODUCTIONConserving biodiversity in tropical landscapes is a major challenge to scientists andconservationists (Tilman et al 2011) The tropics sustain most of the worldrsquos describedbiodiversity (Brown 2014) but suffer from the highest rates of deforestation and land useintensificationmdashmainly due to rapid agricultural expansion (Wright amp Muller-Landau2006 Gibbs et al 2010) This scenario yields mosaics of multiple artificialsemi-naturalareas of land use abruptly (fragmentation) or gradually (variegation) bordering naturalhabitats (Fischer amp Lindenmayer 2007) often resulting in the loss of native species (Arroyo-Rodriguez et al 2013) and in some cases local extinctions (Newmark 1991 Lehtinen ampRamanamanjato 2006) Such a depletion of biodiversity can result in biotic homogenization(Solar et al 2015) and consequently alter ecosystem functioning leading to deteriorationin the provisioning of ecosystem goods and services (Olden et al 2004 Clavel Julliard ampDevictor 2011Mitchell et al 2015)

Nevertheless human modified landscapes are still useful for species conservation(Chazdon et al 2009 Gardner et al 2009) especially if patches of natural vegetation arepresent (Nichols et al 2007 Fahrig 2013) and if the matrix is highly suitable for localbiodiversity (Prugh et al 2008 Franklin amp Lindenmayer 2009) Consequences of humanactivities may differ between fragmented and variegated landscapes While fragmentedlandscapes generally host isolated populations in habitat patches surrounded by hostilematrices (Fahrig 2003 Fahrig et al 2011) variegated landscapes present multiple artificialor semi-natural land uses that are gradually different from the natural habitats (McIntyreamp Hobbs 1999 Fischer amp Lindenmayer 2007) These landscapes may be more permeable tospecies movement exhibiting distinct biodiversity patterns in response to human activities(Daily 2001 Roumls Escobar amp Halffter 2012) and thus have high conservation value (Barlowet al 2010 Gibson et al 2011)

Nevertheless studies on the responses of biodiversity to tropical variegated landscapeswith large number of replicates are scarce (Fischer amp Lindenmayer 2007) The few relatedstudies indicate that variegated landscapes are in fact more connected than fragmentedlandscapes and have variable importance for biodiversity conservation sustaining highlevels of biodiversity in the tropics (Barlow et al 2007 Roumls Escobar amp Halffter 2012)However these studies are generally conducted in recently modified areas that are undergreat influence from natural habitats (Barlow et al 2010) Thus it is difficult to disentanglethe contribution ofmodified areas from that of natural habitats to biodiversity conservationThe investigation of older tropical variegated landscapes may offer important insightsabout future biodiversity patterns in these areas Also considering that most of the worldrsquosterrestrial ecosystems and one quarter of worldrsquos threatened species are living outsideprotected areas understanding the importance of these human modified landscapes forbiodiversity conservation becomes crucial (Rodrigues et al 2004 Jenkins amp Joppa 2009Troupin amp Carmel 2014 Ekroos et al 2016)

Here we investigated biological communities present in tropical variegated landscapesthat have been subject to intense pressures of urbanization agriculture and livestockproduction since the 18th century (ca 300 years) (Zemella 1990 Vilela 2007) Our

Costa et al (2017) PeerJ DOI 107717peerj3125 222

studied area is composed of mosaics of semi-deciduous secondary forest fragments (ofvariable sizes and regeneration status) native and introduced pasturelands monocultures(mainly coffee plantations) and hedgerows (forest corridors) (Burel 1996 Oliveira-Filho amp Fluminhan-Filho 1999 Castro amp Van den Berg 2013) We aimed to assess dungbeetle communities in twelve 300-year-old tropical variegated landscapes in order to findempirical evidence of their conservation value to biodiversity We used dung beetles as ourfocal taxa because species of this group are abundant in our studied area easily sampledand identified play important ecological roles are associated with vertebrates and arewidely used as bioindicators (Nichols et al 2007 Nichols et al 2008 Nichols amp Gardner2011 Gillett et al 2016) Furthermore dung beetle communities from tropical forests aregreatly influenced by vegetation structure due to their association with specific climatic(physiological intolerance) and edaphic conditions (Halffter amp Arellano 2002 see Nicholsamp Gardner 2011 and references therein Griffiths et al 2015)

We investigated the extent towhich dungbeetle species richness abundance andbiomassand community structure are affected by (1) land use and cover classmdashLUCC (ie forestcorridors forest fragments coffee plantations and pastures) and (2) structural differencesamong habitats (ie variation in canopy cover local vegetation heterogeneity and soil sandcontent) We also assessed the importance of landscape variegation to conservation of dungbeetle regional diversity (3) disentangling the relative contribution of nestedness-resultantand spatial turnover to beta-diversity patterns in variegated landscapes

MATERIAL amp METHODSThe studywas carried out in a 70-km2 area of themunicipality of Lavras southeastern Brazil(2115primeSndash2118prime25

primeprime

S 4500prime57primeprime

Wndash4454prime34primeprime

W) in the transition between two biodiversityhotspots the Cerrado (tropical savanna) and the Atlantic Forest (semideciduous seasonalforest) (Fig 1) The climate in this region is humid subtropical (Cwa) according toKoumlppen climate classification and experiences cold-dry winters and hot-rainy summersThe annual precipitation and mean temperature are 1460 mm and 204 C respectivelyand the elevation varies between 967 m and 1055 m (Schiffler 2003 Dantas Carvalho ampFerreira 2007) During the studied period (January 2011mdashsummer) the total precipitationand mean temperature were about 1364 mm and 230 C respectively (Source INMETnetwork data) This season is considered the best period of the year to sample dung beetlesin tropical areas (Martiacutenez amp Vasquez 1995 Lobo amp Halffter 2000Milhomem VazdeMelloamp Diniz 2003)

The studied area has historically experienced pressures from agro-pastoral activitiesand urbanization which generated the variegated landscapes Overall the landscapesare composed of fragments of secondary semideciduous forests (forest fragments)interconnected by hedgerows (forest corridors) and human settlements immersed inmatrices of coffee plantations or introduced pastures We delimited twelve 500 times 500m experimental landscapes in which we selected one site of each of the main landuses forest fragments (with average size of 1825 ha) forest corridors (colonization ofland plot boundary ditches typical of this regionmdashCastro amp Van den Berg 2013) coffee

Costa et al (2017) PeerJ DOI 107717peerj3125 322

Figure 1 Study area map showing (A) localization of the studied region within the Minas Gerais StatemdashBrazil (B) the 12 studied landscapes (represented by each sample point) and the different types of land useand cover classes in the studied region Map thematic source Tainaacute Assis

plantations (traditional managementmdashCoffea arabica L) and pastures grazed by cattle(exotic plantsmdashUrochloa spp) The selected forest corridors were connected to forestfragments and adjacent to coffee plantations and pastures Coffee plantations were notpresent in four of the landscapes Thus we sampled 12 forest fragments 12 forest corridors12 pastures and eight coffee plantations We established one 150 m transect in each of the44 sampling sites

Sampling of dung beetlesWe sampled dung beetles using six pitfall traps 30 m apart from each other along eachtransect (Total 6 times 44 = 264 traps) We used a smaller distance between traps thanrecommended by some authors (Larsen amp Forsyth 2005 Silva amp Hernaacutendez 2015) becauseour fragments were small and our sample unit was the area of LUCCs (each pitfall value ofeach dung beetle attribute was pooled in a sample unique per transect) Each pitfall trapconsisted of a plastic container (19 cm diameter 11 cm depth) half filled with a solutionof saline and detergent (5) to break the surface tension of the water and preserve dungbeetles and a hanging bait compartment with a lid to protect against rain and desiccationby the sun We placed the pitfall traps which were baited with 30 g of homogenized humanfeces between 900 am and 400 pm and left them open for a 48 h period After samplingdung beetles were sorted counted and identified to the lowest taxonomic level possiblewith the help of available taxonomic keys (eg Vaz-de-Mello et al 2011) and the CREN

Costa et al (2017) PeerJ DOI 107717peerj3125 422

(Neotropical Dung Beetles Reference Collection at the Universidade Federal of Lavras)Voucher specimens were deposited at CREN All dung beetles were dried at 40 C in orderto preserve the specimens and to obtain their dry body weight In order to calculate speciesmean biomass we weighed 20 individuals (or the maximum possible) of each species usinga precision scale balance (00001 g)

Dung beetles were sampled on farms with the permission of the landholders We alsopossessed an IBAMASISBIO license (number 10061-1) in the name of Julio LouzadaIn addition the feces used in the study was donated by the authors who all agreed ondonating it

Measuring structural differences among land classesTo explain variation in the studied community parameters we measured threeenvironmental variables reflecting structural differences among LUCCs canopy cover(CC) local vegetation heterogeneity (LH) and soil sand content (SS) To estimate CC weused hemispherical canopy photographs taken 15 m above the soil next to each pitfalltrap with an 8-mm fisheye lens (Engelbrecht amp Herz 2001) We analyzed the photographsusing the software Gap Light Analyzer 20 (GLA Frazer Canham amp Lertzman 1999) andquantified the percentage of pixels related to vegetation in each photograph as a proxyfor canopy cover We measured the fractal dimension (number that characterizes thegeometry of a fractal) of the understory vegetation to use as a proxy for LH To do sowe took photographs of the understory according to a methodology adapted from Nobis(2005) and analyzed them in the software SIDELOOK (Nobis 2005) which calculates thefractal dimension of each photograph Photographs of a black panel (1 times 1 m) placedbehind vegetation 3 m away were taken with a camera with a 52-mm lens positioned 1m above the soil adjacent to each pitfall trap To measure SS we used a homogenizationof all the soil samples taken next to the pitfall traps of a transect (Total = 44 samples)Homogenized soil samples were analyzed for their texture meaning content of sand siltand clay in each soil sample As these variables are highly correlated we only used sandcontent (percentage in the sample) as a measurement of soil structure Sand content is asoil variable related to an important dung beetle behavior (digging) that plays an essentialrole in ecosystem functioning (Halffter amp Edmonds 1982Davis 1996Griffiths et al 2015)

Data analysisComparisons of species richness among LUCCs could be biased because of possibledifferences in sample coverage or low sample coveragemdashwhich would mean that dungbeetle communities were under-sampled (Chao amp Jost 2012) To make more accuratecomparisons we calculated LUCC-level sampling coverage using iNEXT package in R(Chao et al 2014 Hsieh Ma amp Chao 2016) This package also allows us to comparespecies richness of standardized samples at the same sample completeness based on ararefactionextrapolation sampling curve (RE curve) (Hsieh et al 2016)

We used dung beetle species richness abundance biomass and community structure asresponse variables and LUCC as the explanatory variable to answer our first question Firstwe used Generalized Linear Models (GLM) with species richness (total number of species

Costa et al (2017) PeerJ DOI 107717peerj3125 522

per transect) abundance (total number of individuals per transect) and biomass (total drybody weight per transect) as response variables We submitted models to pairwise contrastanalysis (lsmeans packagemdashLenth 2016) in order to combine statistically similar classesof land uses and cover Models were built and compared using R language (R DevelopmentCore Team 2015) Second we conducted Principal Coordinate Analysis (PCOmdashGower1966) followed by Permutational Multivariate Analysis of Variance (PERMANOVAmdashMcArdle amp Anderson 2001)mdashto test for significant clustering of sites with respect todifferent LUCCsWe used community structure (matrix based on square-root transformedabundance data and Bray Curtis dissimilarity index) as the response variable Finally weperformed tests for homogeneity of multivariate dispersions (PERMDISPmdashAnderson2006) to check for differences in variance dispersion of community structure data amongLUCCs This analysis was performed using the software Primer v6 with PERMANOVA +(Clarke amp Gorley 2006)

We used dung beetle community structure as the response variable and CC LH andSS as explanatory variables to answer our second question First we used HierarchicalPartitioning to assess the influence of CC LH and SS on species richness abundance andbiomass This method provides an estimate of the independent effects of each explanatoryvariable on the response variable (Chevan amp Sutherland 1991 Mac Nally 2000) Weperformed this analysis using R language (R Development Core Team 2015) Secondwe used Distance-based Multivariate Analysis for a Linear Model (DistLM Legendre ampAnderson 1999 McArdle amp Anderson 2001) to assess the influence of CC LH and SS oncommunity structure DistLM analyzes and models the relationship between a multivariatedata cloud and one or more independent variables (Anderson Gorley amp Clarke 2008)DistLM allows independent variables to be fitted individually or together in user specifiedsets The DistLM routine was based on the AICc model selection criterion (Burnham ampAnderson 2004) using a lsquolsquostep-wisersquorsquo selection procedure Primer 60 and PERMANOVA+for PRIMER software were used (Clarke amp Gorley 2006 Anderson Gorley amp Clarke 2008)

In order to answer our third question we decomposed beta diversity of dung beetlecommunities into spatial turnover and nestedness-resultant components to determinetheir relative contributions to beta-diversity patterns in the studied landscapes Thebeta diversity was decomposed into Soslashrensen (βSOR) and Simpson (βSIM) dissimilarityindices (Baselga 2010) Soslashrensen (βSOR) dissimilarity represents the total beta diversity andincorporates both species replacement and nestedness-resultant dissimilarities Simpson(βSIM) dissimilarity describes species turnover or replacement and it is equal to βSOR inthe absence of nestedness Thus the difference between these indices is a measure of thenestedness-resultant component of beta diversity (βSNE= βSORminusβSIM (Baselga 2010))

We calculated multiple-site dissimilarity to estimate the overall beta diversity of dungbeetle communities among all sites in each landscape (Baselga 2013) In order to representthe relative contribution of the nestedness-resultant component to overall beta diversitywe calculated its proportion for overall multiple-site dissimilarity (βratio = βSNEβSOR)Where βratio lt05 represents dominance of species replacement in beta diversity patternsand βratio gt05 represents dominance of the nestedness-resultant component (Dobrovolskiet al 2012)

Costa et al (2017) PeerJ DOI 107717peerj3125 622

02 04 06 08 10

0

10

20

30

40

Coffee plantationForest corridorForest fragment

Pasture

Sample coverage

Spe

cies

div

ersi

ty

FF FC CP P

A B

Figure 2 Sample coverage-based species accumulation curve of dung beetle sampled in forest frag-ment forest corridor coffee plantation and pasture of 12 landscapes in Lavras Brazil (A) Estimatedaverage species richness and standard deviation at the same sample coverage (776) in FF forest frag-ment FC forest corridor CP coffee plantation and P pasture (B) The shaded area indicates the 95confidence interval and the dashed line represents extrapolation data

RESULTSWe collected a total of 2695 individuals of 52 species of dung beetles from the tribesAteuchini (three genera 11 species) Delthochilini (five genera 13 species) Coprini (fivegenera 14 species) Oniticellini (one genus four species) Onthophagini (one genustwo species) and Phanaeini (four genera eight species) Of these 28 species occurred inforest fragments (1549 individuals) forest corridors (603 individuals) and pastures (211individuals) and 19 species in coffee plantations (332 individuals) (Table 1) The highestaverage sample coverage of our sampled LUCCs was for forest fragment samples (SC =938) and the lowest coverage was in pasture samples (SC = 776mdashcoffee plantation= 9249 and forest corridor = 8556) (Table S1) When all LUCCs were comparedat equal sample coverage (in this case we used rarefied coverages at the lowest averagevaluemdashapp 776) estimated average species richness showed a different trend than thoseof the raw data All LUCCs had the same estimated species richness (Fig 2B Table S2)Our RE coverage-based curves (based on pooled data) showed similar patterns of speciesaccumulation between forest fragments and forest corridors (Fig 2A)

We observed significant differences in the mean species richness abundance andbiomass among LUCCs (Frichness = 28978 p = 004 df = 3 Fabundance = 67067plt 0001 df = 3 Fbiomass= 71122 plt 0001 df = 3) Overall forest fragments exhibitedhigher values than the other LUCCs while forest corridors and coffee plantations

Costa et al (2017) PeerJ DOI 107717peerj3125 722

Table 1 Dung beetles collected at Forest fragments (FF) Forest corridors (FC) Coffee plantation (CP)and Pasture (P) in LavrasmdashBrazil

TribeSpecies FF FC CP P Biome

AteuchiniAteuchus aff carbonarius (Harold 1868) 0 1 0 0 ndashAteuchus sp 6 1 0 0 ndashAteuchus striatulus (Borre 1886) 0 0 0 2 CCanthidium aff sulcatum (Perty 1830) 0 0 2 0 ndashCanthidium aterrimumHarold 1867 403 28 125 19 AFCanthidium barbacenicum Borre 1886 0 3 2 4 CCanthidium decoratum (Perty 1830) 0 0 0 5 CCanthidium sp1 0 1 2 0 ndashCanthidium sp2 0 0 0 1 ndashCanthidium sp3 2 0 0 0 ndashUroxys sp 1 3 14 1 ndashDelthochiliniCanthon sp1 13 67 0 0 ndashCanthon aff podagricusHarold 1868 0 0 0 10 ndashCanthon chalybaeus Blanchard 1843 0 2 48 0 AFCanthon lituratus (Germar 1824) 0 0 0 2 CCanthon septemmaculatus histrio (Serville 1828) 18 0 1 0 CCanthon sp2 1 1 0 0 ndashCanthon virensMannerheim 1829 0 0 0 5 CDeltochilum orbignyi (Blanchard 1845) 0 0 0 1 ndashDeltochilum rubripenne Gory 1831 81 1 0 0 ndashDeltochilum sp 9 4 0 0 ndashPseudocanthon aff xanthurus (Blanchard 1843) 0 0 0 5 ndashScybalocanthon korasakiae Silva 2011 219 96 0 0 AFSylvicanthon foveiventris Schmidt 1920 390 95 0 0 AFCopriniDichotomius aff rotundigena Felsche 1901 1 3 2 0 ndashDichotomius affinis Felsche 1910 26 4 0 2 ndashDichotomius bicuspis (Germar 1824) 22 34 68 1 AFDichotomius bos (Blanchard 1843) 2 2 4 78 CDichotomius carbonariusMannerheim 1829 5 52 44 7 CAFDichotomius depressicollis (Harold 1867) 4 1 0 0 AFDichotomius fissusHarold 1867 5 0 0 0 AFDichotomius mormon Ljungh 1799 234 29 2 1 AFDichotomius nisus (Olivier 1789) 0 0 0 2 CDichotomius sp 0 2 0 0 ndashEutrichillum hirsutum Boucomont 1928 0 0 0 3 CIsocopris inhiatus (Germar 1824) 0 0 0 2 ndashOntherus aztecaHarold 1869 5 4 0 0 CAFTrichillum externepunctatum Borre 1886 1 0 0 6 C

(continued on next page)

Costa et al (2017) PeerJ DOI 107717peerj3125 822

Table 1 (continued)

TribeSpecies FF FC CP P Biome

OniticelliniEurysternus caribaeus (Herbst 1789) 10 9 0 0 AFEurysternus cyanescens Balthasar 1939 1 0 0 0 CAFEurysternus hirtellus Dalman 1824 38 6 0 0 AFEurysternus parallelus Castelnau 1840 44 146 1 1 CAFOnthophaginiOnthophagus aff hircullusMannerheim 1829 0 0 2 0 ndashOnthophagus ranunculus Arrow 1913 0 0 2 25 CPhanaeiniCoprophanaeus cyanescens Olsoufieff 1924 1 5 4 4 CAFCoprophanaeus horusWaterhouse 1891 0 0 5 11 CCoprophanaeus spitzi (Pessocirca 1935) 0 0 2 4 CDendropaemon sp 1 1 0 0 ndashOxysternon palaemon (Laporte 1840) 0 0 0 6 CPhanaeus kirbyi Vigors 1825 0 0 0 1 CPhanaeus palaeno Blanchard 1843 0 0 2 2 CPhanaeus splendidulus Fabricius 1781 6 2 0 0 AFAbundance 1549 603 332 211 ndashSpecies Richness 28 28 19 28 ndash

NotesAF species registered in Atlantic Forest samples C species registered in Cerrado samples lsquolsquondashrsquorsquo uncertainwithout identifi-cation based on Almeida et al (2011) Campos amp Hernaacutendez (2013) and Audino Louzada amp Comita (2014) Costa et al 2016unpublished data

had intermediate values and pastures the lowest values however some pair-to-paircomparisons were not significantly different (Fig 3) All pair-to-pair comparisons andresults can be accessed in Table S3

The PCO revealed three distinct groups (forest fragment + forest corridor coffeeplantation and pasture) with axis 1 and 2 explaining 436 of the variation in structure(species compositionmdashFig 4)However dung beetle community structurewas significantlydifferent among the LUCCs (PERMANOVA pseudo-F = 80969 p= 0001 df = 3)(Table S4) The LUCCs also exhibited differences in the dispersion of the variance of thecommunity structure data (PERMDISP F = 35964 p= 005 df = 3) with higher valuesin pasture in comparison to forest fragment (t = 29631 p= 0017) and coffee plantation(t = 41819 p= 0003) (Table S4)

Canopy cover (CC) significantly influenced all community parameters studiedHierarchical partitioning revealed positive effects of CC on dung beetle species richness(8307 of independent effect) abundance (801) and biomass (8304) (Table 2)Likewise community structure exhibited the same pattern (2235 of independent effect)(Pseudo-F = 12090 plt 0001 df = 42) (Table 3)

In the variegated landscapes studied the decomposition of beta diversity revealed thatthe main process driving beta diversity in these landscapes was spatial turnover with βratiolt05 in all landscapes (Fig 5)

Costa et al (2017) PeerJ DOI 107717peerj3125 922

Figure 3 Boxplots showing the richness (A) abundance (B) and biomass (C) of dung beetle across theland use and cover classes in LavrasmdashBrazil FF forest fragment FC forest corridor CP coffee planta-tion and P pasture Different letters means significant differences at p lt 005 among the land uses andcover classes

Costa et al (2017) PeerJ DOI 107717peerj3125 1022

Figure 4 PCO biplot of BrayndashCurtis similarity matrix based on square-root transformed dung beetleabundance data in land use and cover classes

Table 2 Results of hierarchical partitioning analyses with all the environmental variables

Variables I I J I + J =R2radicI

(a) Total richnessCC 83072 0127 0009 014 +037SS 5375 0008 minus0003 0005 009LH 11553 0018 0012 003 013

(b) Total abundanceCC 80101 0130 0031 0162 +036SS 0424 00007 minus00007 000003 003LH 19475 0032 0032 0063 018

(c) Total biomassCC 83047 013 0026 0156 +04SS 0537 00008 00006 0001 003LH 16415 0026 0025 0050 02

NotesI percentage of independent effect I independent explanatory power of the variable J Joint explanatory power of the vari-able with all other variables listed ( I + J = R2) univariate squared correlation and

radic(I ) square root of the independent ex-

planatory power which may be interpreted as the independent correlation with the response variable the sign is allocated tothat of the univariate correlation CC canopy cover SS soil sand content LH local vegetation heterogeneityIndicates statistically significant variables at ple 005

Costa et al (2017) PeerJ DOI 107717peerj3125 1122

Figure 5 Results from decomposition of dung beetle beta diversity in four land use and cover classesat 12 variegated landscapes in LavrasmdashMG BrazilGrey bars represent spatial turnover component pro-portion (βSIMβSOR) white bars represent nestedness-resultant component proportion (βSNEβSOR) andblack dots represent overall values of beta diversity (βSNE+βSIM) in each landscape

Table 3 Results of distance based linear models (DistLM) Response variable is dung beetle speciescomposition and predictor variables are canopy cover (CC) soil sand content (SS) and local vegetationheterogeneity (LH)

Variable AICc SS(trace) Pseudo-F P Prop Cumulative resdf

Marginal testsSS ndash 35551 1071 03621 0025 ndash 42CC ndash 31958 121 00001 02235 ndash 42LH ndash 18590 63 00001 0130 ndash 42

Sequential testsCC 34896 31958 121 00001 02235 02235 42

NotesProp Proportion of explained variation

DISCUSSIONThis study evaluated the ecological patterns of dung beetle communities in variegatedtropical landscapes highlighting the importance of these landscapes for conservation oftropical biodiversity Our 12 studied landscapes presented diverse dung beetle communitiesthat were structurally different among the LUCCs (high beta diversity) and capable of

Costa et al (2017) PeerJ DOI 107717peerj3125 1222

sustaining several species from both the Atlantic Forest and Cerrado (mainly from openphysiognomies)

Dung beetle communities respond differently to the LUCCs present in our studiedlandscapes with variation in species richness abundance biomass and communitystructure The presence of well-defined communities in each LUCC highlights theimportance of their maintenance for conserving regional diversity Because LUCCs varyregarding their permeability to forest (fragments corridors and coffee) and Cerrado(coffee and pasture) species the studied landscapes were able to sustain a significantgroup of native species from both the Atlantic Forest and Cerrado found in this region(Table 1) (Schiffler 2003 Almeida amp Louzada 2009 Almeida et al 2011 Barragaacuten et al2011 Gries et al 2012) However in each modified LUCC (pasture coffee plantation andforest corridor) we found more species at the regional level (all landscapes) than at thelocal scale (per landscape) The low dung beetle abundance and biomass observed leadus to believe that these land use classes are used as transitional habitats for dung beetlessuch as ecological corridors or stepping-stones (Fagan Cantrell amp Cosner 1999 Estrada ampCoates-Estrada 2002 Fischer amp Lindenmayer 2007 Diacuteaz Galante amp Favila 2010 Almeidaet al 2011)

These land use classes also harbor exclusive species as the main process promoting betadiversity in these landscapes is spatial replacement In consolidated landscapes such asstudied in this paper environmental filters have already acted in each of the studied LUCCsshowing that some dung beetle species apparently recognize different LUCCs as habitats(Webb et al 2010) The ability of a species to survive in human-modified landscapes is ofgreat importance (Gardner et al 2009) since currently most ecosystems suffer some levelof perturbation Of the environmental factors measured in this study the most importantwas canopy cover This variable is often reported in scientific literature as a proxy forhabitat quality and resource availability for dung beetles (Halffter amp Matthews 1966Halffter 1991 Halffter amp Arellano 2002 Louzada et al 2010 Audino Louzada amp Comita2014) Although soil and vegetation parameters can influence dung beetle communities(Gries et al 2012 Farias et al 2015) because they can affect larvae survival (Osberg Doubeamp Hanrahan 1994 Davis et al 2010) the present work found LH and SS not liable fordetermining dung beetle community structure

Our results provide additional evidence that variegation of a landscape can allow speciesmovement between habitats of variable suitability favoring their long-term persistence(Doerr et al 2014) Such a scenario offers a better outlook for biodiversity conservationthan scenarios resulting from fragmentation Fragmentation tends to confine species inreduced patches of low-quality habitat eventually leading populations to suffer fromproblems related to endogamy (eg reduced genetic variability enhanced susceptibility todiseases and stochastic events) and local extinctions (Keller amp Largiadegraver 2003Keyghobadia2007 Delaney Riley amp Fisher 2010) In our studied landscapes we recorded several speciestypical of the Atlantic Forest in forest fragments forest corridors and coffee plantations(Table 1) In addition Cerrado species were dominant in pastures (Table 1) suggesting thatlocal species may be able to persist in human modified landscapes if enough time is givenand if introduced habitats conserve at least some structural similarity with natural ones

Costa et al (2017) PeerJ DOI 107717peerj3125 1322

Based on the ability of our studied modified habitats to conserve native species andcontribute to increased regional diversity we encourage the consideration of variegation ofpreviously fragmented landscapes in the management of human modified landscapes (RoumlsEscobar amp Halffter 2012) Diversification of LUCCs may benefit biodiversity improveregional heterogeneity and connectivity and help maintain the provisioning of criticalecological functions (Andresen 2002 Nichols et al 2008 Braga et al 2012 Braga et al2013) In transitional areas this diversification could be even more important for speciesconservation Human modification generally favors the occurrence of open landscapesby (a) suppressing the natural habitat of species from at least two habitat types (egin a forest-forest transition) (b) favoring species of a single habitat type (eg forest tonon-forest transition) (c) homogenizing two open habitat types (eg conversion of nativesavannas and fields into pastures) or (d) suppressing at least two natural habitat types (egurbanization)

Finally we caution against ignoring the negative effects of deforestation and habitatdegradation For instance a recent study by Barlow et al (2016) showed that humandisturbances in the Amazon forest were responsible for reducing the diversity of dungbeetles birds and plants with almost two times the strength of deforestation This reinforcesthe irreplaceability of natural habitats for biodiversity conservation and highlights theneed to reduce disturbances in the remaining habitats Our results reflect a 300-yearold scenario of human-induced modification which despite showing relatively goodprospects for biodiversity conservation was responsible for the suppression of large areasof natural habitats Therefore our studied landscapes may have already experienced strongbiodiversity losses In face of current rates of tropical deforestation degradation andland use intensification such landscapes are becoming more common We encouragethe variegation of landscapes as a complementary initiative of current conservationpractices (eg protection of natural habitats and establishment of reserves) Togetherthese management strategies may achieve partial recovery of biodiversity and ecosystemfunctioning even when landscapes are not able to sustain the entire biodiversity of nativecommunities (Barlow et al 2010 Gray et al 2016)

ACKNOWLEDGEMENTSWe thank all the owners of the studied farms for allowing our fieldwork to take place ontheir properties JW Barretto F Frazatildeo M Yankous and F Franca for field assistance andDr F Silva for help with dung beetle identification UFLA provided institutional support

ADDITIONAL INFORMATION AND DECLARATIONS

FundingFAPEMIG CNPq and CAPES supported the research and people involved in this paperThe funders had no role in study design data collection and analysis decision to publishor preparation of the manuscript

Costa et al (2017) PeerJ DOI 107717peerj3125 1422

Grant DisclosuresThe following grant information was disclosed by the authorsFAPEMIGCNPqCAPES

Competing InterestsThe authors declare there are no competing interests

Author Contributionsbull Cristiane Costa conceived and designed the experiments performed the experimentsanalyzed the data wrote the paper prepared figures andor tablesbull Victor Hugo F Oliveira wrote the paperbull Rafaella Maciel conceived and designed the experiments performed the experimentsreviewed drafts of the paperbull Wallace Beiroz performed the experiments prepared figures andor tables revieweddrafts of the paperbull Vanesca Korasaki conceived and designed the experiments performed the experimentsanalyzed the data prepared figures andor tables reviewed drafts of the paperbull Julio Louzada conceived and designed the experiments contributed reagentsmaterial-sanalysis tools reviewed drafts of the paper

Field Study PermissionsThe following information was supplied relating to field study approvals (ie approvingbody and any reference numbers)

The MMAIBAMASISBIO approved this work under licence number 10061-1

Data AvailabilityThe following information was supplied regarding data availability

The raw data has been supplied as Supplementary File

Supplemental InformationSupplemental information for this article can be found online at httpdxdoiorg107717peerj3125supplemental-information

REFERENCESAlmeida SSP Louzada JNC 2009 Estrutura da comunidade de Scarabaeinae (Scarabaei-

dae Coleoptera) em fitofisionomias do Cerrado e sua importacircncia para a con-servacatildeo Neotropical Entomology 3832ndash43 DOI 101590S1519-566X2009000100003

Almeida SSP Louzada JNC Sperber C Barlow J 2011 Subtle land-use change andtropical biodiversity dung beetle communities in cerrado grasslands and exoticpastures Biotropica 43704ndash710 DOI 101111j1744-7429201100751x

Costa et al (2017) PeerJ DOI 107717peerj3125 1522

Andresen E 2002 Dung beetles in a Central Amazonian rainforest and their eco-logical role as secondary seed dispersers Ecological Entomology 27257ndash270DOI 101046j1365-2311200200408x

AndersonMJ 2006 Distance-based tests for homogeneity of multivariate dispersionsBiometrics 62245ndash253 DOI 101111j1541-0420200500440x

AndersonMJ Gorley RN Clarke KR 2008 PERMANOVA+ for PRIMER guide tosoftware and statistical methods Plymouth PRIMER-E 214

Arroyo-Rodriguez V RoumlsM Escobar F Melo FPL Santos BA Tabarelli M ChazdonR 2013 Plant beta-diversity in fragmented rain forests testing floristic homog-enization and differentiation hypotheses Journal of Ecology 1011449ndash1458DOI 1011111365-274512153

Audino LD Louzada J Comita L 2014 Dung beetles as indicators of tropical forestrestoration success is it possible to recover species and functional diversity Biolog-ical Conservation 169248ndash257 DOI 101016jbiocon201311023

Barlow J Gardner TA Araujo IS Aacutevila-Pires TC Bonaldo AB Costa JE Esposito MCFerreira LV Hawes J HernandezMIM HoogmoedMS Leite RN Lo-Man-HungNF Malcolm JR Martins MB Mestre LAMMiranda-Santos R Nunes-Gutjahr ALOveral WL Parry L Peters SL Ribeiro-Junior AB Da Silva MNF Da Silva MottaC Peres CA 2007 Quantifying the biodiversity value of tropical primary secondaryand plantation forests Proceedings of the National Academy of Sciences of the UnitedStates of America 10418555ndash18560 DOI 101073pnas0703333104

Barlow J Lennox GD Ferreira J Berenguer E Lees AC Mac Nally R Thomson JRFerraz SF Louzada J Oliveira VH Parry L Solar RR Vieira IC Aragatildeo LE BegottiRA Braga RF Cardoso TM De oliveira Jr RC Souza CMJR Moura NG NunesSS Siqueira JV Pardini R Silveira JM Vaz-De-Mello FZ Veiga RC VenturieriA Gardner TA 2016 Anthropogenic disturbance in tropical forests can doublebiodiversity loss from deforestation Nature 7535144ndash147

Barlow J Louzada J Parry L HernaacutendezMIM Hawes J Peres CA Vaz-de-MelloFZ Gardner TA 2010 Improving the design and management of forest strips inhuman-dominated tropical landscapes a field test on Amazonian dung beetlesJournal of Applied Ecology 47779ndash788 DOI 101111j1365-2664201001825x

Barragaacuten F Moreno CE Escobar F Halffter G Navarrete D 2011 Negative impactsof human land use on dung beetle functional diversity PLOS ONE 6(3)e17976DOI 101371journalpone0017976

Baselga A 2010 Partitioning the turnover and nestedness components of beta diversityGlobal Ecology and Biogeography 19134ndash143 DOI 101111j1466-8238200900490x

Baselga A 2013 Separating the two components of abundance-based dissimilaritybalanced changes in abundance vs abundance gradientsMethods in Ecology andEvolution 4552ndash557 DOI 1011112041-210X12029

Braga RF Korasaki V Andresen E Louzada J 2013 Dung beetle community andfunctions along a habitat-disturbance gradient in the Amazon a rapid assessmentof ecological functions associated to biodiversity PLOS ONE 8(2)e57786DOI 101371journalpone0057786

Costa et al (2017) PeerJ DOI 107717peerj3125 1622

Braga RF Korasaki V Audino LD Louzada J 2012 Are dung beetles driving dung-flyabundance in traditional agricultural areas in the Amazon Ecosystems 151173ndash1181DOI 101007s10021-012-9576-5

Brown JH 2014Why are there so many species in the tropics Journal of Biogeography418ndash22 DOI 101111jbi12228

Burel F 1996Hedgerows and their role in agricultural landscapes Critical Reviews inPlant Sciences 15169ndash190 DOI 10108007352689199610393185

BurnhamKP Anderson DR 2004Multimodel inference understanding AIC and BIC inmodel selection Sociological Methods amp Research 33261ndash304DOI 1011770049124104268644

Campos RC HernaacutendezMIM 2013 Dung beetle assemblages (Coleoptera Scarabaeinae)in Atlantic forest fragments in Southern Brazil Rivista Brasileira de Entomologia5747ndash54 DOI 101590S0085-56262013000100008

Castro GC Van den Berg E 2013 Structure and conservation value of high-diversityhedgerows in southeastern Brazil Biodiversity and Conservation 222041ndash2056DOI 101007s10531-013-0524-2

Chao A Gotelli NJ Hsieh TC Sander EL Ma KH Colwell RK Ellison AM 2014Rarefaction and extrapolation with Hill numbers a framework for sampling andestimation in species diversity studies Ecological Monographs 84(1)45ndash67DOI 10189013-01331

Chao A Jost L 2012 Coverage-based rarefaction and extrapolation standardizing sam-ples by completeness rather than size Ecology 932533ndash2547 DOI 10189011-19521

Chazdon RL Peres CA Dent D Sheil D Lugo AE LambD Stork NE Miller SE 2009The potential for species conservation in tropical secondary forests ConservationBiology 231406ndash1417 DOI 101111j1523-1739200901338x

Chevan A SutherlandM 1991Hierarchical partitioning The American Statistician4590ndash96 DOI 1023072684366

Clarke KR Gorley RN 2006 Primer v6 user manualtutorial Plymouth PlymouthMarine Laboratory

Clavel J Julliard R Devictor V 2011Worldwide decline of specialist species towarda global functional homogenization Frontiers in Ecology and the Environment9222ndash228 DOI 101890080216

Daily GC 2001 Ecological forecasts Nature 411245 DOI 10103835077178Dantas AAA Carvalho LG Ferreira E 2007 Classificacatildeo e tendecircncias climaacuteticas em

Lavras MG Ciecircncias Agroteacutecnicas 311862ndash1866Davis ALV 1996 Seasonal dung beetle activity and dung dispersal in selected South

African habitats implications for pasture improvement in Australia AgricultureEcosystems and Environment 58157ndash169 DOI 1010160167-8809(96)01030-4

Davis AL Scholtz CH Kryger U Deschodt CM StruumlmpherWP 2010 Dung beetleassemblage structure in Tswalu Kalahari Reserve responses to a mosaic of land-scape types vegetation communities and dung types Environmental Entomology39811ndash820 DOI 101603EN09256

Costa et al (2017) PeerJ DOI 107717peerj3125 1722

Delaney KS Riley SPD Fisher RN 2010 A rapid strong and convergent genetic re-sponse to urban habitat fragmentation in four divergent and widespread vertebratesPLOS ONE 5(9)e12767 DOI 101371journalpone0012767

Diacuteaz A Galante E Favila ME 2010 The effect of the landscape matrix on the distribu-tion of dung and carrion beetles in a fragmented tropical rain forest Journal of InsectScience 1081 DOI 1016730310108101

Dobrovolski R Melo AS Cassemiro FAS Diniz-filho JAF 2012 Climatic historyand dispersal ability explain the relative importance of turnover and nestednesscomponents of beta-diversity Global Ecology and Biogeography 21(2)191ndash197DOI 101111j1466-8238201100671x

Doerr ED Doerr VAJ Davies MJ McGinness HM 2014 Does structural connectivityfacilitate movement of native species in Australiarsquos fragmented landscapes Asystematic review protocol Environmental Evidence 39 DOI 1011862047-2382-3-9

Ekroos J Oumldman AM Andersson GKS Birkhofer K Herbertsson L Klatt BK OlssonO Olsson PA Persson AS Prentice HC Rundloumlf M Henrik smith G 2016 Sparingland for biodiversity at multiple spatial scales Frontiers in Ecology and Evolution3145 DOI 103389fevo201500145

Engelbrecht BMJ Herz HM 2001 Evaluation of different methods to estimate under-storey light conditions in tropical forests Journal of Tropical Ecology 17207ndash224DOI 101017S0266467401001146

Estrada A Coates-Estrada R 2002 Dung beetles in continuous forest forest fragmentsand in an agricultural mosaic habitat island at Los Tuxtlas Mexico BiodiversityConservation 111903ndash1918 DOI 101023A1020896928578

FaganWF Cantrell RS Cosner C 1999How habitats edges changes species interac-tions The American Naturalist 153165ndash182 DOI 101086303162

Fahrig L 2003 Effects of habitat fragmentation on biodiversity Annual Review ofEcology Evolution and Systematics 34487ndash515DOI 101146annurevecolsys34011802132419

Fahrig L 2013 Rethinking patch size and isolation effects the habitat amount hypothe-sis Journal of Biogeography 401649ndash1663 DOI 101111jbi12130

Fahrig L Baudry J Brotons L Burel FG Crist TO Fuller RJ Sirami C SiriwardenaGMMartin J-L 2011 Functional landscape heterogeneity and animal biodiversityin agricultural landscapes Ecology Letters 14101ndash112DOI 101111j1461-0248201001559x

Farias PM Arellano L HernaacutendezMIM Ortiz SL 2015 Response of the copro-necrophagous beetle (Coleoptera Scarabaeinae) assemblage to a range of soilcharacteristics and livestock management in a tropical landscape Journal of InsectConservation 19947ndash960 DOI 101007s10841-015-9812-3

Fischer J Lindenmayer DB 2007 Landscape modification and habitat fragmentation asynthesis Global Ecology and Biogeography 16265ndash280DOI 101111j1466-8238200700287x

Costa et al (2017) PeerJ DOI 107717peerj3125 1822

Franklin JF Lindenmayer DB 2009 Importance of matrix habitats in maintainingbiological diversity Proceeding National Academic Science of the United States ofAmerica 106349ndash350 DOI 101073pnas0812016105

Frazer GW Canham CD Lertzman KP 1999 Gap light analyzer (GLA) Version 20New York Simon Fraser University and the Institute of Ecossistem Studies 36p

Gardner TA Barlow J Chazdon R Ewers RM Harvey CA Peres CA Sodhi NS 2009Prospects for tropical forest biodiversity in a human-modified world Ecology Letters12561ndash582 DOI 101111j1461-0248200901294x

Gibbs HK Ruesch AS AchardMK ClaytonMK Holmgren P Ramankutty N FoleyA 2010 Tropical forests were the primary sources of new agricultural land in the1980s and 1990s Proceedings of the National Academy of Sciences of the United Statesof America 10716732ndash16737 DOI 101073pnas0910275107

Gibson L Lee TM Koh LP Brook BW Gardner TA Barlow J Peres CA Bradshaw CJLauranceWF Lovejoy TE Sodhi NS 2011 Primary forests are irreplaceable forsustaining tropical biodiversity Nature 478378ndash381 DOI 101038nature10425

Gillett CPDT Johnson AJ Barr I Hulcr J 2016Metagenomic sequencing of dungbeetle intestinal contents directly detects and identifies mammalian fauna BioRxivpreprint DOI 101101074849

Gower JS 1966 Some distance properties of latent root and vector methods used inmultivariate analysis Biometrika 53325ndash338 DOI 101093biomet533-4325

Gray CL Simmons BI Fayle TMMann DJ Slade EM 2016 Are riparian forest reservessources of invertebrate biodiversity spillover and associated ecosystem functions inoil palm landscapes Biological Conservation 194176ndash183DOI 101016jbiocon201512017

Gries R Louzada J Almeida S Macedo R Barlow J 2012 Evaluating the impacts andconservation value of exotic and native tree afforestation in Cerrado grasslands usingdung beetles Insect Conservation and Diversity 5175ndash185DOI 101111j1752-4598201100145x

Griffiths HM Louzada J Bardgett RD BeirozW Franca F Tregidgo D Barlow J 2015Biodiversity and environmental context predict dung beetle-mediated seed dispersalin a tropical forest field experiment Ecology 961607ndash1619 DOI 10189014-12111

Halffter G 1991Historical and ecological factors determining the geographical dis-tribution of beetles (Coleoptera Scarabaeidae Scarabaeinae) Folia EntomologicaMexicana 82195ndash238

Halffter G Arellano L 2002 Response of dung beetle diversity to human-inducedchanges in a tropical landscape Biotropica 34(1)144ndash154DOI 101111j1744-74292002tb00250x

Halffter G EdmondsWD 1982 The nesting behavior of dung beetles (Scarabaeinae)mdashanecological and evolutive approach Mexico Instituto de Ecologiacutea MAB 242p

Halffter G Matthews EG 1966 The natural history of dung beetles of the subfamilyScarabaeinae (Coleoptera Scarabaeidae) Folia Entomologica Mexicana 121ndash312

Costa et al (2017) PeerJ DOI 107717peerj3125 1922

Hsieh TC Ma KH Chao A 2016 iNEXT an R package for interpolation and ex-trapolation of species diversity (Hill numbers)Methods in Ecology and Evolution7(12)1451ndash1456 DOI 1011112041-210X12613

Jenkins CN Joppa L 2009 Expansion of the global terrestrial protected area systemBiological Conservation 1422166ndash 2174 DOI 101016jbiocon200904016

Keller I Largiadegraver CR 2003 Recent habitat fragmentation caused by major roads leadsto reduction of gene flow and loss of genetic variability in ground beetles Proceedingsof the Royal Society B 270(1513)417ndash423 DOI 101098rspb20022247

Keyghobadia N 2007 The genetic implications of habitat fragmentation for animalsCanadian Journal of Zoology 85(10)1049ndash1064 DOI 101139Z07-095

Larsen TH Forsyth A 2005 Trap spacing and transect design for dung beetle biodiver-sity studies Biotropica 37322ndash325 DOI 101111j1744-7429200500042x

Legendre P AndersonMJ 1999 Distance-based redundancy analysis testing multi-species responses in multifactorial ecological experiments Ecological Monographs691ndash24 DOI 1018900012-9615(1999)069[0001DBRATM]20CO2

Lehtinen RM Ramanamanjato J-B 2006 Effects of rainforest fragmentation and cor-relates of local extinction in a herpetofauna from Madagascar Applied Herpetology395ndash110 DOI 101163157075406776984248

Lenth RV 2016 Least-squares means the R package lsmeans Journal of StatisticalSoftware 69(1)1ndash33 DOI 1018637jssv069i01

Lobo JM Halffter G 2000 Biogeographical and ecological factors affecting the altitu-dinal variation of mountainous communities of coprophagous beetles (ColeopteraScarabaeoidea) a comparative study Annals of the Entomological Society of America93115ndash126 DOI 1016030013-8746(2000)093[0115BAEFAT]20CO2

Louzada J Lima AP Matavelli R Zambaldi L Barlow J 2010 Community structureof dung beetles in Amazonian savannas role of fire disturbance vegetation andlandscape structure Landscape Ecology 25631ndash641 DOI 101007s10980-010-9448-3

Mac Nally R 2000 Regression and model building in conservation biology biogeog-raphy and ecology the distinction between and reconciliation of lsquopredictiversquo andlsquoexplanatoryrsquo models Biodiversity and Conservation 9655ndash671DOI 101023A1008985925162

Martiacutenez IM Vasquez AA 1995 Influencia de algunos factores ambientales sobre lareproducion em Canthon cyanellus cyanellus Le Conte (Coleoptera ScarabaeidaeScarabaeinae) Elytron 95ndash13

McArdle BH AndersonMJ 2001 Fitting multivariate models to community data acomment on distance-based redundancy analysis Ecology 82290ndash297DOI 1018900012-9658(2001)082[0290FMMTCD]20CO2

McIntyre S Hobbs R 1999 A framework for conceptualizing human effects on land-scapes and its relevance to management and research models Conservation Biology131282ndash1292 DOI 101046j1523-1739199997509x

MilhomemMS Vaz-de-Mello FE Diniz IR 2003 Teacutecnicas de coleta de besouroscopronecroacutefagos no cerrado Pesquisa Agropecuaacuteria Brasileira 381249ndash1256

Costa et al (2017) PeerJ DOI 107717peerj3125 2022

Mitchell MGE Suarez-Castro AF Martinez-HarmsMMaronMMcalpine C GastonKJ Johansen K Rhodes JR 2015 Reframing landscape fragmentationrsquos effects onecosystem services Trends in Ecology and Evolution 30190ndash198DOI 101016jtree201501011

NewmarkWD 1991 Tropical forest fragmentation and the local extinction of under-story birds in the eastern Usambara Mountains Tanzania Conservation Biology567ndash78 DOI 101111j1523-17391991tb00389x

Nichols E Larsen T Spector S Davis AL Escobar F Favila M Vulinec K TheScarabaeinae Research Network 2007 Global dung beetle response to tropicalforest modification and fragmentation a quantitative literature review and meta-analysis Biological Conservation 1371ndash19 DOI 101016jbiocon200701023

Nichols E Spector S Louzada J Larsen T Amezquitad S Favila ME 2008 Ecologicalfunctions and ecosystem services provided by Scarabaeinae dung beetles BiologicalConservation 1411461ndash1474 DOI 101016jbiocon200804011

Nichols ES Gardner TA 2011 Dung beetles as a candidate study taxon in appliedbiodiversity conservation research In Simmons LW Ridsdill-Smith TJ eds Ecologyand evolution of dung beetles New Jersey Wiley-Blackwell 267ndash291

Nobis M 2005 SideLook 11mdashImaging software for the analysis of vegetation structurewith true-colour photographs Disponiacutevel em Available at httpwwwapplecoch

Olden JD Leroy Poff N Douglas MR Douglas ME Fausch KD 2004 Ecological andevolutionary consequences of biotic homogenization Trends in Ecology amp Evolution1918ndash24 DOI 101016jtree200309010

Oliveira-Filho AT Fluminhan-FilhoM 1999 Ecologia da vegetacatildeo do Parque FlorestalQuedas do Rio Bonito Revista Cerne 5051ndash064

Osberg DC Doube BM Hanrahan SA 1994Habitat specificity in African dungbeetles the effect of soil type on the survival of dung beetle immatures (ColeopteraScarabaeidae) Tropical Zoology 71ndash10 DOI 10108003946975199410539236

Prugh LR Hodges KE Sinclair ARE Brashares JS 2008 Effect of habitat area and isola-tion on fragmented animal populations Proceedings of the National Academy of Sci-ences of the United States of America 10520770ndash20775 DOI 101073pnas0806080105

RDevelopment Core Team 2015 R a language and environment for statisticalcomputing Vienna R Foundation for Statistical Computing Available at httpwwwr-projectorg

Rodrigues ASL Andelman SJ Bakarr MI Boitani L Brooks TM Cowling RM FishpoolLDC Da Fonseca GAB Gaston KJ HoffmannM Long JS Marquet PA Pilgrim JDPressey RL Schipper J Sechrest W Stuart SN Underhill LGWaller RWWattsMEJ Yan X 2004 Effectiveness of the global protected area network in representingspecies diversity Nature 428(2004)640ndash643 DOI 101038nature02422

RoumlsM Escobar F Halffter G 2012How dung beetles respond to a human-modifiedvariegated landscape in Mexican cloud forest a study of biodiversity integratingecological and biogeographical perspectives Diversity and Distributions 18377ndash384DOI 101111j1472-4642201100834x

Costa et al (2017) PeerJ DOI 107717peerj3125 2122

Schiffler G 2003 Fatores determinantes da riqueza local de espeacutecies de Scarabaeidae(Insecta Coleoptera) em fragmentos de Floresta Estacional Semideciacutedua In Con-clusion study of Masterrsquos degree (dissertation)ndashMestrado em AgronomiaConcentracatildeoem Entomologia Lavras Universidade Federal de Lavras 68 p

Silva PG HernaacutendezMIM 2015 Spatial patterns of movement of dung beetle species ina tropical forest suggest a new trap spacing for dung beetle biodiversity studies PLOSONE 10e0126112 DOI 101371journalpone0126112

Solar RRDC De Barlow J Ferreira J Berenguer E Lees AC Thomson JR Louzada JMaueacutes M Moura NG Oliveira VHF Chaul JCM Schoereder JH Vieira ICG MacNally R Gardner TA 2015How pervasive is biotic homogenization in human-modified tropical forest landscapes Ecology Letters 181108ndash1118DOI 101111ele12494

Tilman D Balzer C Hill J Befort BL 2011 Global food demand and the sustainableintensification of agriculture Proceedings of the National Academy of Sciences of theUnited States of America 10820260ndash20264 DOI 101073pnas1116437108

Troupin D Carmel Y 2014 Can agro-ecosystems efficiently complement protected areanetworks Biological Conservation 169158ndash166 DOI 101016jbiocon201311009

Vaz-de-Mello FZ EdmondsWD Ocampo F Schoolmeesters P 2011 A multilingualkey to the genera and subgenera of the subfamily Scarabaeinae of the NewWorldZootaxa 28541ndash73

Vilela MS 2007 A formacatildeo histoacuterica do campo de SantrsquoAna das Lavras do Funil LavrasEditora Indi 450 p

Webb CT Hoeting JA Ames GM PyneMI Poff NL 2010 A structured and dynamicframework to advance traits-based theory and prediction in ecology Ecology Letters13267ndash283 DOI 101111j1461-0248201001444x

Wright SJ Muller-Landau HC 2006 The uncertain future of tropical forest speciesBiotropica 38443ndash445 DOI 101111j1744-7429200600177x

Zemella MP 1990O abastecimento da capitania das minas gerais no seacuteculo XVIII 2edition Satildeo Paulo Hucitec-Edusp 247

Costa et al (2017) PeerJ DOI 107717peerj3125 2222