Natural enemies depend on remnant habitat size in agricultural landscapes

Natural enemies depend on remnant habitat size in agricultural landscapes

Mainara Xavier Jordani?E′rica Hasui?Vin?′cius Xavier da Silva

In recent decades,the consequences of habitat fragmentation have been of growing concern,because it is particularly important to understand how fragmentation may affect biodiversity,an ecological service.We tested two hypotheses:(1)that natural fragment size in agricultural landscapes indirectly affects the herbivore through effects on natural predator populations;and(2)predator activity into the crop reduces along the distance from the natural fragment edge.From 2008 and 2009,we conducted our study in seven forest remnants and in surrounding coffee plantations(fragments ranged from 6 to 105 ha, mean 49.28±36.60 ha)in Southern Minas Gerais,Brazil. Birds were sampled by point counts,and insect predation was evaluated by using an artificial insect model(Koh and Menge 2006).Our results suggest that although there weremany potential predators(e.g.,wasps,ants,birds,and mammals),birds were the most important taxon unit.The covariance analysis supported the hypothesis that patch size affected the numberof larvae predation by overalltaxi, but there was no support for a distance effect.These findings suggest that natural enemies’ecological service (mainly from birds)declined with remnant reduction, which has implications not only for human welfare,but also in strengthening the economic justifications for conserving the remaining natural habitats and biodiversity in agricultural landscapes.

Ecosystem service·Fragmentation· Isolation·Insectivore·Predation

Introduction

Increasing human population size and food consumption per person have resulted in an expansion in agricultural landscapes and a concurrentreduction of naturalhabitats to smaller,isolated fragmented patches.These changes in land use are the major drivers of biodiversity loss,because these patches become too small to support particular species or too far apart to ensure regional persistence of metapopulation dynamics(Tilman etal.2001;Fahrig 2003; Tscharntke et al.2005).

Although most people give little thought to how dependent they are on biodiversity,crucial services for humanity are disappearing or becoming inefficient(Daily 1997;Tscharntke et al.2005;Whelan et al.2008).If conserved and managed appropriately,biodiversity can contribute to agricultural productivity and sustainability of ecosystem services(such as pest control,crop pollination, soil fertility,protection against soil erosion in waterways,and the removal of excessive nutrients).Natural pest regulation is an important service with economic and human health benefits.

Pests,particularly herbivorous insects,destroy 37 percent of the potential crop yield(Pimentel et al.1997). Despite millions of tons of synthetic pesticides used annually,farmers frequently failed to combatpests because many have evolved resistance to them.To avoid immunity to specific chemicals,farmers need to maintain as many weapons in the pest control arsenal as possible and alternate these strategies(Sodhi and Ehrlich 2010).However, these chemical compounds can also kill beneficial species (such as pollinators,or natural predators of the pest)and also put human health at risk through food and water contamination.

Achieving success in natural pest control depends on many factors and can easily be disrupted(Bianchi et al. 2006).There is increasing evidence that landscape structure and composition can interfere with this service (Tscharntke et al.2005;Bianchi et al.2008).Several studies have shown that species richness and abundance of natural enemies are affected by the proximity of a natural habitat patch(Tscharntke et al.2007).In this case,predators cross natural habitat edges into neighboring crops, where they significantly reduce prey populations(Cronin 2003).Thus,managing action on natural habitats in agricultural ecosystems assumes that significant predator incursions will across natural habitat edges and result in enhanced biological control(Denno et al.2005).

The question of the optimal size and distance of natural habitats to enhance natural enemies,and consequently maximize ecosystem service,is not well understood.For example,Puckett et al.(2009)showed higher foraging habits of insectivorous birds within 20 m of the edge,but other studies also found a foraging distance pattern at least 1 km apartfrom a naturalhabitat(Jonsson et al.2010).The distance pattern depends on the biology of individual species,mainly related to habitat specialization and dispersion ability in the crop.

We test the hypothesis of incremental predator activity into the crop along natural patch-size gradients.From a natural enemies’perspective,increasing natural patch size can result in higher abundance and diversity of food resource(MacArthur and Wilson 1967;Connor et al. 2000).If predator populations exhibit positive response to food availability,then higher predator abundance and greater predator effects are likely to be found in larger patches(With 2002;Langellotto and Denno 2004).This increment in predator abundance can be propagated into the crops,if predators can disperse freely across natural habitat edges(Cronin 2003).

Alternatively,the opposite hypothesis of the relationship between patch size and foraging activity can be expected due edge-related phenomena.In this case,in small natural patches with high edge-to-area ratio,prey mortality will increase as a result of predator incursion into the crops (Cronin et al.2004).There are data to support both scenarios of increasing and decreasing impacts across a patch gradient(Denno et al.2005).In this study,we tested two alternative hypotheses and also evaluate the distance effect from the naturalhabitat patch in the agriculturallandscape.

Materials and methods

Study area

Fieldwork was conducted in fragments in the southern Brazilian state of Minas Gerais,Brazil(21°25′48.03′′S, 45°56′51.76′′W,Fig.1).Altitude in this area averages 880 m asl,with a mean annual temperature of 23°C and annual rainfall of 1,413 mm(Costa 1998).The original vegetation is classified as seasonal semideciduous forest (Oliveira Filho and Fontes 2000).However,now the landscape is highly fragmented,with only four percent of native forest remaining in various successional stages (Fundac?a?o SOS Mata Atla?ntica and INPE 2009).The landscape matrix is mainly composed of pastures and coffee and sugar cane plantations.

The landscape analysis was conducted using satellite images(CBERS),with a 20 m resolution,using ArcGIS 9.2TMsoftware to visually classify mature forest remnants in a 30 km radius of the municipality of Alfenas,Brazil. The patch size was then calculated to select seven forest remnants in a size gradient ranging from 6 to 105.9 ha (mean 49.28±36.60 ha,Fig.1).The matrix surrounding forest remnants was composed by coffee plantation.

Fig.1 Location ofthe study area in the State of Minas Gerais,Brazil, showing the location of the sample sites(dark gray)

Bird sampling

In each sample site,six points were chosen:two points inside the fragment(located at 100 m from the nearest forest edge),two in the forest edge(located at one meter from the nearest forest edge)and two inside the coffee plantation(located at 100 m from forest edge).Each point was at least 100 m apart from the other(Fig.2).Each fragment was sampled three times in the wet season in the years of 2008 and 2009,always in the morning.

Birds were sampled using the point count method with limited distance(30 m,Develey 2003).The time for sampling in each point was 10 min.All birds were recorded,but only insectivorous(including omnivores)understory birds were considered in this study,because they are potential predators of caterpillars.The classification of foraging behaviour and vertical stratification followed Willis(1979),Fitzpatrick(1980),Ridgely and Tudor (1994,1997),Remsen and Robinson(1990),Stotz et al. (1996),Sick(2001)and Gomes et al.(2008).The taxonomic classification of the bird species followed CBRO (Comite?Brasileiro de Registros Ornitolo′gicos 2011).

Caterpillar predation

An experimental approach was used to compare the number of caterpillars found in the fragments.Using modeling clay,an oil-based and nontoxic compound,artificial caterpillars were constructed to capture the bite marks of potentialpredators.The artificialcaterpillars had a standard size(30 mm×7 mm),shape(cylindrical)and color (green).Due to malleability of model and the type of impression upon its retrieval,the predator taxa can be identified using a method successfully employed in other studies(Freitas and Oliveira 1996;Koh and Menge 2006).

Fig.2 Sampling unit design of seven sites in Minas Gerais,Brazil. Each sample site was composed of six pointcounts(white circles)and five transects(white lines).White circles representthe locations where birds were sampled by point counts and the white transects where predation models were distributed

The artificialcaterpillarswere randomly distributed along five parallel trails(20 caterpillars per trail,length=5 m) per fragment:two inside the fragment,two inside the coffee plantation,and one on the edge forest(Fig.2).Each trailwas atleast25 m from the other.Each caterpillarwas secured on the top of leaves at 1.5 m above ground using plastic glue. Afterfive days,the caterpillars were collected and examined under a stereomicroscope in orderto determine the nature of the bite marks inflicted by potential predators(e.g.,mammals,birds,ants and wasp;Fig.3).

Data analysis

The relative importance of predators was determined using a two-way ANOVA with habitat and predator type as the categoricalpredictors.A subsequent Tukey’s test was done to assess where there was a difference.In this analysis we excluded the edge-collected data.

Through an analysis of covariance(ANCOVA),three models were established,representing all possible n-way combinations of the two predictor variables:patch size and predator type.Later these models were compared and selected by Akaike information criterion(AIC—Akaike Information Criterion)and then adjusted for small samples (AICc)(Burnham and Anderson 2002).This highlighted the most plausible models to explain the variation in the number of caterpillars.The Akaike criterion is defined by likelihood AICc=-2+2*K*(n/(nk-1)),where maximum likelihood is the likelihood of the data,K is the number of parameters in the model,and n is the size of sample.This generates a ranking of the best to worst model,and AICc differences≥2 are considered great,and the best support is given to the model with lowest AICc.

The plausibility of the models was evaluated through the values ofΔAIC(difference between a determined and the lowest value of AIC among all models tested)and wi, which is the Akaike weight,which varies between 0 and 1, and estimates the weight of evidence in favor of a model i given the set of models compared.

To assess whether the distance from the fragmentaffects the amount of caterpillars attacked by predators,we also conducted an analysis of covariance to determine the best model relating the number of attacked caterpillars to the distance from fragmentand type of predator.So we explore the gradient interior-edge-coffee plantation.Our study did nothave a controlarea for absolute lack of forestfragments with more than 500 hectares within 100 km radius.

Furthermore,a Spearman correlation was used to evaluate the relationship between the abundance index of insectivorous bird and the number of artificial caterpillars predated by birds.

Fig.3 Examples of different bite marks found on artificial caterpillars:a mammals,b birds and c ants

Results

Relative importance of predators

Birds,insects,and mammals were the main predators of caterpillars in the study site.After five days,they attacked 21.7–60.3%of the artificial caterpillars.After excluding mammal predations from the statistical analysis due to an insufficient sample size(only two occurrences),there were significant differences between predators(i.e.,birds,ants, and wasps).The birds preferentially attacked in the forest (two-way ANOVA:F2,46=4.1;p=0.02),but even so, their attacks in the coffee plantation were higher than those from wasps and ants(Tukey test:p＜0.05)(Fig.4).

Fig.4 Numberofartificiallarvae attacked by birds,wasp and ants in the interiorofthe forestfragments and in the coffee plantation.Points represent mean±SE

Interaction between predator type and patch size effects

The best-supported model constructed in the analysis of covariance(GLM–ANCOVA)indicated thatthe predations were affected only by the combination of predator type and patch size(Table 1).There was no support for the individual prediction models(ΔAICc≥2).Specifically,this model showed a positive relationship between patch size and number of predation,but the intensity of those attacks was conditioned by the predator type(Fig.5a).Birds attacked more with an increase in patch size.The same thing happened with ants and wasps,but at lower levels.

Distance from remnants effect on predations in the coffee plantations

There was no support for the effect of distance from remnants according to the analysis of covariance(GLM–ANCOVA).The most plausible model included only the predator type(Table 2).The contribution of distance as predictor variable was very low(predictor weight=0.03), mainly when wasp and ant predations were considered (Fig.5b).

Potential bird predators

From a total of 23 insectivorous understory bird species in the sample sites,five may be potential candidates for pestpredators,because they mainly attack prey on foliage and stems and also occur in the coffee plantations(Table 3).In particular,Basileuterus leucoblepharus(Spearman correlation:rs=0.82 and p=0.02)and Dysithamnus mentalis (Thamnophilidae,Spearman correlation:rs=0.89 and p=0.006)showed a significant correlation between the number of caterpillars attacked and the number of bird contacts.

Table 1 GLM–ANCOVAmodelresults compared with the nullmodel, ranked by Akaike’s Information Criterion corrected for small sample size(AICc),predicting the numberofartificialcaterpillarpredations

Fig.5 Numberofartificiallarvae attacked by birds,wasps and ants in differentforestfragmentsizes(a)and differentdistance from the edge of forestfragments(b)

Table 2 GLM–ANCOVA model results compared with the null model,ranked by Akaike’s Information Criterion corrected for small sample size(AICc),predicting the number of artificial caterpillar predations

Discussion

Influence of natural habitat patch on pest regulation service

The economic reasons for saving and restoring natural habitats are growing increasingly influential for politicians and resource managers.It is of utmost importance to identify strategies that generate positive co-benefits for production,biodiversity,and local people.This experimentalstudy found evidence thatthe size of natural habitat patches in the agriculturallandscape interfered with natural predators.The predator service increases with the increasing size of the adjacentnaturalhabitat,being able to reduce larvae by 21.7–60.3%after five days.

These percentages may be higher since we studied predation in small-sized patches(ranging from 6 to 105.9 ha),and therefore higher percentages may be expected in larger natural patches.For example,Greenberg et al.(2000)found that birds reduced the abundance of large arthropods(＞5 mm)in a coffee plantation by 64–80%and also with lower levels of leaf damage. Borkhataria et al.(2006)expanded the insect size classes, demonstrating thatbirds can also reduce small-sized insect populations in a coffee plantation.Furthermore,pest control services in the crops were not influenced by the distance from natural habitats,at least 100 m apart from the edge.

Potential pest predators

Due to the study design,we only have indirect evidence about the type of predators.Our results indicated that predation intensity is species specific,where bird species were more effective control agents than ants or wasps, although birds were more affected by patch-size variation. Differential responses among taxa may be due to differences in their biological traits,such as ecological specialization,matrix use,and organism dispersal capacity(Henle et al.2004).

However,interactions of species traits and landscape structure must be considered.These predators aredependent on resources not contained within a single habitat type and are likely influenced by the landscape structure of all required patch-habitat types.The ability to use resources in different habitats is dependent on biological traits of species and is influenced both by the characteristics of the patch(food supply,predation risk, competition pressure)and the characteristics of landscape (habitat complexity,diversity,quality,and patchiness). This spatial association of required habitat patches influenced the distribution patterns of predation across the cropnatural habitat interface(Tscharntke et al.2005).

Table 3 Bird species recorded in point count observations of seven forest remnants and surrounding coffee matrix in Alfenas,MG

Our results suggest,at least two distribution patterns of predator attacks.The first pattern,represented by ant and wasp predators,showed homogeneous distribution of attacks in the crop-naturalhabitatinterface withoutspecific preference for one habitat type or the other.The second pattern,described for bird predators,showed heterogeneous distribution of attacks with higher attacks in the natural habitats than in the crops.

Five insectivorous bird species occurred in the matrix and are possible candidates of these attacks.However, Basileuterus leucoblepharus and Dysithamnus mentalis had a strong positive correlation between the number of larvae attacks and the number of individual bird contacts. Thus,they were potentially the most important pest predator in the study area.The positive effect of bird attacks with the naturalpatch size was explained by the addition of bird species of higher sensitivity to human disturbances (sensus Stotz et al.1996),such as Hemitriccus orbitatus and Drymophila ferruginea,which are restricted to larger fragments.This work supports an earlier suggestion that insectivorous birds were important in the reduction of herbivorous insect pests(Altegrim 1989;Strong et al. 2000)or plant damages in both natural habitats and coffee plantation(Greenberg et al.2000;Borkhataria et al.2006; Kellermann etal.2008;Van Baeletal.2008;Johnson etal. 2009).For example,they are able to reduce the population of the forest pest Lepidoptera,which the birds feed on,by 20–100%(Barbaro and Battisti 2011).

Despite the lower level of attacks by wasps and ants, previous studies have demonstrated the importance of these taxa as natural enemies of coffee plantation pests such asleafminer,Leucoptera coffeellum(Gue′rin-Me′neville) (Lepidoptera:Lyonetiidae),which causes great economic losses in some New World coffee-producing countries, such as Brazil,Colombia,Cuba,Guatemala,Peru,and Puerto Rico(Parra 1985;Reis et al.2000;Fragoso et al. 2001).A study in Brazil on population dynamics of this species demonstrated that wasps could reduce pest populations by 69%(Reis et al.2000).A collection of wasp predator species from prior studies documented,at least, six genera and 11 species—including Protonectarina sylverae Saussure,Polybia paulista Ihering,Polybia occidentalis Olivier,Polybia scutellaris White,Brachygastra lecheguana Latreille,and Synoeca surinama cyanea Linnaeus—thatfeed on coffee leafminers(Fragoso etal.2001; Lomeli-Flores 2009).However,the findings of Reis et al. (2000)contrast to Lomeli-Flores(2009),who found that ants were the principal coffee leafminer predators in Mexico.At least 12 species were observed preying on eggs,larvae,or pupae,from the Camponotus,Pseudomyrmex,and Azteca genera.Lomeli-Flores(2009)suggests that the differences in the relative importance of predator species between countries could be due to differences in coffee-farm microclimatic conditions and/or management regimes.Brazilian coffee production is mainly under sunny conditions and farms are intensively managed,while in Mexico,production is under shady conditions and farms are traditionally managed,but this difference needs to be further tested and verified.

Natural habitat influence on predation in crops

Potential mechanism behind patch size

The positive relationship between the amount of predation in crops and the size of adjacent forest is consistent with the resource-concentration hypothesis(Root 1973).This hypothesis conjectures that population density should be positively correlated with patch area.Previous work found evidence for this relationship in bird and insect species (Connor et al.2000).Several potential mechanisms,such as demographic effect,may be responsible for this pattern, and differences among species(Connor et al.2000).In the demographic effect,smaller patches should have lower density,due to greater vulnerability to stochastic conditions or the higher probability of an Allee effect(Vergara and Hahn 2009).Therefore,a denser population in larger patches could increase prey consumption in resource-rich habitats(Bianchi and Wa¨ckers 2008).

Distance effect from remnant

Based on theoretical predictions(Strong 1992;Polis and Strong 1996),we would expect a stronger impact of predators on herbivorous prey near natural habitats.However,our results did not support this prediction for all taxa. The absence of the distance effect may be due to the fine spatial scale used in the experimental designs,considering the high mobility of the natural enemies in the crops. Future analyses thatincorporate larger distance ranges may influence or change the observed relationships.Klein et al. (2006)found a significant distance effect in wasps,working with distances as far as 1,400 m to the nearest natural forest.

Conservation and management implications

Arthropod pests have been estimated to cause around 14% losses in the gross domestic product(GDP)in developed countries and 38 percentin developing countries(Zambolim etal.2008).For example,considering thatthe GDP of Brazilian agribusiness reached R$822.9 billion in 2011 (NationalConfederation of Agriculture and Cattle Breeding of Brazil),an estimated R$312.7 billion was lostin the year due to arthropod pests.Chemical pesticides are still the dominant form of control for many of these pests,but they also imply additionalcosts in the form of human health and degraded environment.On the other hand,the management and conservation ofnaturalhabitatscan increase production, withoutthese negative additionalcosts,due the conservation and enhancement of natural enemies(Jonsson et al.2010). These methods can provide naturalenemies with a favorable microclimate condition and place forshelter,dormancy,and alternative food sources(Landis etal.2000).Severalstudies showed thatdensity and diversity ofnaturalenemies tend to be higher in landscapes with a high proportion of non-crop vegetation(Bianchietal.2006).Generally,the percentage of habitatarea in a given landscape hasa strong correlation with the mean patch size and the size of largest patch(Fahrig 2003).Therefore,landscapes with higher percentages of habitat areas often correspond to landscapes to where patches are large.

Our studies support these previous predictions,showing a positive relationship between the size of natural habitat patches and the amount of predation predators.However, some questions about this relationship and other landscape ecology issues remain.For example,what is the optimal size of a naturalhabitat to provide natural pest control and, the same time,to maximize the crop production?Should landscape restoration efforts focus on enlarging existing natural patches or building new patches?Will restored or managed landscape really restore the natural enemy’s composition and consequently the ecological process? What is the minimum distance between natural patches to facilitate natural enemy’s species migration in a managed landscape?What is the optimal spatial scale to which the predation process mainly responds?

Future research addressing these questions can facilitate biological conservation associated with crop production increases.Concluding,our results suggest that the natural enemies’service(mainly from birds)declined with natural habitat reduction in the agricultural landscape.This has implications not only for human welfare,but also in strengthening the economic justifications for conserving remaining natural habitats.

AcknowledgmentsWe wish to express ourgratitude to M.Raniero, M.F.V.Silva,E.Pessoni,and other members of the Laborato′rio de Ecologia de Fragmentos Florestais(ECOFRAG),who have been valuable friends,assisted in fieldwork,and made essential comments at different phases of this study.This manuscript also greatly benefited from the comments of Alberto Jose′Arab Olavarrieta.We also thank several private landowners who permitted access to their properties.Universidade Federal de Alfenas provided logistical support.We received financial support from Fundac?a?o de Amparo a` Pesquisa do Estado de Minas Gerais FAPEMIG-VALE S/A(Process #RDP-00104-10)and Conselho Nacional de Desenvolvimento Cient?′fico e Tecnolo′gico(CNPq)(Process#472250/2010).We appreciated the improvements in English language made by Jim Hesson of http://www.AcademicEnglishSolutions.com

Altegrim O(1989)Exclusion of birds from bilberry stands:impacton insect larval density and damage to the bilberry.Oecologia 79:136–139

Barbaro L,Battisti A(2011)Birds as predators ofthe pine processionary moth(Lepidoptera:Notodontidae).BiolControl56(2):107–114

Bianchi FJJA,Wa¨ckers FL(2008)Effects of flowerattractiveness and nectar availability in field margins on biological control by parasitoids.Biol Control 46:400–408

Bianchi FJJA,Booij CJH,Tscharntke T(2006)Sustainable pest regulation in agricultural landscapes:a review on landscape composition,biodiversity and natural pest control.Proc R Soc Biol Sci Ser B 273(1595):1715–1727

Bianchi FJJA,Goedhart PW,Baveco JM(2008)Enhanced pest controlin cabbage crops near forestin The Netherlands.Landsc Ecol 23(5):595–602

Borkhataria RR,Collazo JA,Groom MJ(2006)Additive effects of vertebrate predators on insects in a Puerto Rican coffee plantation.Ecol Appl 16(2):696–703

Burnham KP,Anderson DR(2002)A practical information-theoretic approach,Second edn.Springer,New York,p 353

Comite?Brasileiro de Registros Ornitolo′gicos—CBRO(2011)Lista das aves do Brasil.10th edn.CBRO.http://www.cbro.org.br. Accessed 15 Nov 2011

Connor EF,Courtney AC,Yoder JM(2000)Individuals–area relationships:the relationship between animalpopulation density and area.Ecology 81(3):734–748

Costa CMR(1998)Biodiversidade em Minas Gerais:um atlas para a sua conservac?a?o.Fundac?a?o Biodiversitas,Belo Horizonte,p 222

Cronin JT(2003)Matrix heterogeneity and host-parasitoid interactions in space.Ecology 84:1506–1516

Cronin JT,Haynes KJ,Dillemuth F(2004)Spider effects on planthopper mortality,dispersal and spatial population dynamics.Ecology 85:2134–2143

Daily GC(1997)Introduction:whatare ecosystem services?In:Daily G(ed)Nature’s services:societal dependence on natural ecosystems.Island Press,Washington,D.C,pp 1–10

Denno RF,Finke DL,Langellotto GA(2005)Direct and indirect effects of vegetation structure and habitat complexity on predator–prey and predator–predator interactions.In:Barbosa P,Castellanos I(eds)Ecology of predator–prey interactions. Oxford University Press,Oxford,pp 211–239

Develey P(2003)Me′todos com estudos com aves.In:Cullen L Jr, Rudran R,Valladares-Pa′dua C(eds)Me′todos de estudos em biologia da conservac?a?o e manejo da vida silvestre.Editora Universidade Federal do Parana′,Curitiba,pp 153–179

Fahrig L(2003)Effects of habitat fragmentation on biodiversity. Annu Rev Ecol Syst 34:487–515

Fitzpatrick JW(1980)Foraging behavior of neotropical tyrant flycatchers.Condor 82(1):43–57

Fragoso DB,Jusselino-Filho P,Guedes RNC,Proque R(2001) Seletividade de inseticidas a vespas predadoras de Leucoptera coffeella(Gue′r.-Me`nev.)(Lepidoptera:Lyonetiidae).Neotrop Entomol 30(1):139–143

Freitas AVL,Oliveira PS(1996)Ants as selective agents on herbivore biology:effects on the behaviour of a non-myrmecophilous butterfly.J Anim Ecol 65:205–210

Fundac?a?o SOS Mata Atla?ntica&(INPE)Instituto Nacional de Pesquisas Espaciais(2009)Atlas dos remanescentes florestais da Mata Atla?ntica,per?′odo de 2005–2008.Fundac?a?o SOS Mata Atla?ntica&Sa?o Jose dos Campos,INPE,Sa?o Paulo

Gomes VSM,Loiselle BA,Alves MAS(2008)Birds foraging for fruits and insects in shrubby restinga vegetation,southeastern Brazil.Biota Neotrop 8(4):21–31

Greenberg R,Bichier P,Angon AC,MacVean C,Perez R,Cano E (2000)The impact of avian insectivory on arthropods and leaf damage in some guatemalan coffee plantations.Ecology 81(6):1750–1755

Henle K,Davies KF,Kleyer M,Margules C,Settele J(2004) Predictors of species sensitivity to fragmentation.Biodivers Conserv 13:207–251

Johnson MD,Levy NJ,Kellermann JL,Robinson DE(2009)Effects of shade and bird exclusion on arthropods and leaf damage on coffee farms in Jamaica’s Blue Mountains.Agrofor Syst 76(1):139–148

Jonsson M,Wratten SD,Landis DA,Tompkins J-ML,Cullen R (2010)Habitat manipulation to mitigate the impacts of invasive arthropod pests.Biol Invasions 12(9):2933–2945

Kellermann JL,Johnson MD,Stercho AM,Hackett SC(2008) Ecologicaland economic services provided by birds on Jamaican Blue Mountain coffee farms.Conserv Biol 22(5):1177–1185

Klein A-M,Steffan-Dewenter I,Tscharntke T(2006)Rain forest promotes trophic interactions and diversity of trap-nesting Hymenoptera in adjacent agroforestry.J Anim Ecol 75(2):315–323

Koh LP,Menge DNL(2006)Rapid assessment of Lepidoptera predation rates in neotropical forest fragments.Biotropica 38(1):132–134

Landis DA,Wratten SD,Gurr GM(2000)Habitat management to conserve naturalenemies of arthropod pests in agriculture.Annu Rev Entomol45:175–201

Langellotto GA,Denno RF(2004)Responses of invertebrate natural enemies to complex-structured habitats:a meta-analytical synthesis.Oecologia 139:1–10

Lomeli-Flores JR(2009)Naturalenemies and mortality factors ofthe coffee leafminer Leucoptera coffeella(Guerin-Meneville)(Lepidoptera:Lyonetiidae)in Chiapas,Me′xico.PhD Dissertation, Texas A&M University,Texas,p 203

MacArthur RH,Wilson EO(1967)The theory of island biogeography.Princeton University Press,Princeton,p 203

Oliveira Filho AT,Fontes MAL(2000)Patterns of floristic differentiation among Atlantic Forests in Southeastern Brazil and the influence of climate.Biotropica 32:793–810

Parra J(1985)Biologia comparada de Leucoptera coffeella(Gue′rin-Me′neville,1842)(Lepidoptera:Lyonetiidae)visando o seu zoneamento ecolo′gico no Estado de Sa?o Paulo.Rev Bras Entomol 29:45–76

Pimentel D,Wilson C,McCullum C,Huang R,Dwen P,Flack J,Tran Q,Saltman T,Cliff B(1997)Economic and environmental benefits of biodiversity.Bioscience 47:747–757

Polis GA,Strong DR(1996)Food web complexity and community dynamics.Am Nat 147:813–846

Puckett HL,Brandle JR,Johnson RJ,Blankenship EE(2009)Avian foraging patterns in crop field edges adjacent to woody habitat. Agric Ecosyst Environ 131(1–2):9–15

Reis R Jr,Lima ER,Vilela EF,Evaldo F,Barros RS(2000)Method for maintenance of coffee leaves in vitro for mass rearing of Leucoptera coffeellum(Gue′rin-Me′neville)(Lepidoptera:Lyonetiidae).An Soc Entomol Bras 29(4):849–854

Remsen JV,Robinson SK(1990)A classification scheme forforaging behavior of birds in terrestrial habitats.Stud Avian Biol 13:144–160

Ridgely RS,Tudor G(1994)The birds of South America:volume 1: the suboscine passerines.University of Texas Press,Austin, p 598

Ridgely RS,Tudor G(1997)The birds of South America:volume 2: the oscine passerines.University of Texas Press,Austin,p 940

Root RB(1973)Organization of a plant-arthropod association in simple and diverse habitats:the fauna of collards(Brassica oleraceae).Ecol Monogr 45:95–120

Sick H(2001)Ornitologia brasileira.Nova Fronteira,Rio de Janeiro, p 912

Sodhi NS,Ehrlich PR(2010)Conservation biology conservation biology for all.Oxford University Press,Oxford,p 344

Stotz DF,Fitzpatrick JW,Parker TA III,Moskovitz DK(1996) Neotropical birds:ecology and conservation.University of Chicago Press,Chicago,p 502

Strong DR(1992)Are trophic cascades all wet?Differentiation and donor-control in species ecosystems.Ecology 73:747–754

Strong DR,Sherry TW,Holmes RT(2000)Bird predation on herbivorous insects:indirect effects on sugar maple saplings. Oecologia 125:370–379

Tilman D,Reich PB,Knops J,Wedin D,Mielke T,Lehman C(2001) Diversity and productivity in a long-term grassland experiment. Science 294:843–845

Tscharntke T,Rand TA,Bianchi FJJA(2005)The landscape context of trophic interactions:insectspillover across the crop–noncrop interface.Ann Zool Fenn 42:421–432

Tscharntke T,Bommarco R,Clough Y,CristTO,Kleijn D,Rand TA, Tylianakis JM,Nouhuys SV,Vidal S(2007)Conservation biologicalcontroland enemy diversity on a landscape scale.Biol Control 43:294–309

Van Bael SA,Philpott SM,Greenberg R,Bichier P,Barber NA, Mooney KA,Gruner DS(2008)Birds as predators in tropical agroforestry systems.Ecology 89(4):928–934

Vergara PM,Hahn I(2009)Linking edge effects and patch size effects:importance of matrix nest predators.Ecol Model 220(9–10):1189–1196

Whelan CJ,Wenny DG,Marquis RJ(2008)Ecosystem services provided by birds.Ann N Y Acad Sci 1134:25–60

Willis EO(1979)The composition of avian communities in remanescent woodlots in southern Brazil.Pap Avulsos Zool (Sao Paulo)33(1):1–25

With KA(2002)Threshold effects of landscape structure on biological control in agroecosystems.Ecol Appl 13(1):314–365

Zambolim L,Conceic?a?o MZD,Santiago T(2008)O que engenheiros agro?nomos devem saber para orientar o uso de produtos fitossanita′rios,3a edn.Universidade Federal de Vic?osa,ANDEF, Vic?osa,p 464

13 April 2013/Accepted:21 June 2013/Published online:29 January 2015

?Northeast Forestry University and Springer-Verlag Berlin Heidelberg 2015

Projectfunding:This work is financially supported from Fundac?a?o de Amparo a`Pesquisa do Estado de Minas Gerais FAPEMIG-VALE S/A (Process#RDP-00104-10)and Conselho Nacional de Desenvolvimento Cient?′fico e Tecnolo′gico(CNPq)(Process# 472250/2010).

The online version is available at http://www.springerlink.com

Corresponding editor:Chai Ruihai

M.X.Jordani

Laborato′rio de Ecologia Animal,Departamento de Zoologia e Bota?nica,Universidade Estadual Paulista-UNESP,Rua Cristo′va?o Colombo,2265,Jardim Nazareth, Sa?o Jose′Do Rio Preto,SP CEP 15054-000,Brazil

E′.Hasui(?)·V.X.da Silva

Laborato′rio de Ecologia de Fragmentos do Sul de Minas Gerais–ECOFRAG,Instituto Cie?ncias da Natureza, Universidade Federal de Alfenas-UNIFAL–MG,Rua Gabriel Monteiro da Silva,700,Centro, Alfenas,MG CEP 37130-000,Brazil e-mail:ericahasui@yahoo.com