Landscape structure shapes the diversity of beneficial insects in coffee producing landscapes

Highlights

• Landscape structure moderates species compositions of beneficial insects in farmland.

• Forest loss in the surrounding landscape resulted in wasp and bee species loss.

• Variations in landscape structure resulted in flower fly species turnover.

• Biodiversity conservation practices must focus on forest conservation and restoration.

Abstract

The expansion of monocultures and the overuse of agrochemicals have resulted in the loss of beneficial insects and disruption of ecosystem services such as pollination and biological control in agricultural landscapes. Bees, wasps and flower flies were our model groups to investigate how landscape structure attributes affect alpha and beta diversity of different beneficial insect groups in Brazilian landscapes containing coffee crops. Species richness and abundance of wasps, and bee richness were positively correlated with forest cover at multiple spatial extents. Bee abundance, and species richness and abundance of flower flies did not respond to any landscape predictor. The community composition of wasps and bees in landscapes with low forest cover was composed of subsets of the communities located in forested landscapes, leading to species loss in structurally impoverished landscapes. High variations in landscape diversity and edge density between landscapes resulted in flower fly species replacement suggesting that pairs of landscapes with high and low diversity of habitat types and edge density harbor different species. Such results indicate that initiatives for the conservation of beneficial insects in the Atlantic Forest biodiversity hotspot must focus on forest conservation and restoration, because high levels of forest loss can result in the loss of wasp and bee species with potential negative consequences for the provision of pollination and pest control services in agroecosystems. Our findings can aid conservationists and policy makers to define priority actions for biodiversity conservation as well as the selection of appropriate spatial scales in landscape planning and management.

Introduction

Balancing sustainable food production and biodiversity conservation is one of the key global environmental challenges (Johnson et al., 2017). Agricultural intensification is considered one of the main threats to biodiversity conservation (Butchart et al., 2010) and the main cause of the decline of natural enemies and pollinators (hereafter beneficial insects), which provide important ecosystem services to both natural and anthropogenic ecosystems (Klein et al., 2007; Geiger et al., 2010; Potts et al., 2010; Ollerton, 2017; Grab et al., 2018). Ecosystem services provided by beneficial insects afford approximately US$ 71.3 billion (2018 dollars, considering inflation) annually in the United States (Losey and Vaughan, 2006). Animal pollination services are responsible for 30% of global food production (Klein et al., 2007) contributing US$235–577 billion to global crop output (Potts et al., 2016), whereas biological control of crop pests has been valued at US$619/ha (2018 dollars) globally (Costanza et al., 1997).

Agricultural intensification encompasses the overuse of agrochemicals and the conversion of natural and semi-natural habitats (e.g., grasslands, forests, hedgerows, abandoned pastures) into expansive monocultures – a process known as landscape simplification (Meehan et al., 2011). Pesticide applications heighten both mortality rates and sublethal effects on insect physiology and behavior such as foraging, fecundity, sex ratio, and learning performance (Stapel et al., 2000; Desneux et al., 2007; Geiger et al., 2010). Landscape simplification negatively affects the capacity of a landscape to provide spatio-temporal insurance through landscape complementation and supplementation (Ouin et al., 2004; Tscharntke et al., 2012; Fahrig, 2017). According to Tscharntke et al. (2012), landscape complementation means that organisms must forage in different habitats to gather spatially separated resources that are necessary to fulfill their needs. For landscape supplementation, landscapes provide organisms with supplemental non-crop and managed habitats with high concentrations of the same type of required resources.

Landscape simplification reduces the amount and diversity of land cover types (compositional heterogeneity) including natural and semi-natural habitats (hereafter non-crop habitats) that play a key role for the maintenance of biodiversity in farmland, particularly invertebrates (Landis et al., 2000; Bianchi et al., 2006). Non-crop habitats are less affected by pesticides and provide refuge and resources for beneficial insects, especially during disturbances in crops such as pesticide application, tillage and harvest operations (Altieri, 1999; Landis et al., 2000; Bianchi et al., 2006; Tscharntke et al., 2012). Moreover, non-crop habitats support high plant diversity, thus providing diverse and continuously available food resources from floral resources (Danner et al., 2016) to alternative preys (Landis et al., 2008). Landscape simplification also alters the spatial arrangement of crop and non-crop habitats in the landscape (configurational heterogeneity) resulting in low connectivity and spill-over opportunities between complementary habitats (Tscharntke et al., 2012; Gámez-Virués et al., 2015). Therefore, habitat loss and fragmentation reduce compositional and configurational landscape heterogeneity leading to the biotic homogenization, species loss and deterioration of ecosystem services (Tscharntke et al., 2012; Gámez-Virués et al., 2015).

The land sharing/land sparing dichotomy has stimulated debate on how to attenuate agricultural intensification in order to conciliate food production and conservation, but sharing/sparing strategies are not mutually exclusive and should work synergistically to avoid undesirable consequences for biodiversity (Fischer et al., 2014; Kremen, 2015). Only a combination of large protected areas (land sparing) surrounded by agroecological wildlife-friendly matrix (land sharing) can improve landscape heterogeneity resulting in high recolonization rates and recovery of degraded ecosystem functioning (Tscharntke et al., 2012; Kremen, 2015).

Studies from temperate and tropical regions indicate that agricultural landscapes with high compositional and configurational heterogeneity can support greater diversity of beneficial insects than do simpler landscapes dominated by monocultures (Meyer et al., 2009; Mandelik et al., 2012; Kennedy et al., 2013; González et al., 2015, González et al., 2016; Kratschmer et al., 2018). However, only a few studies have been conducted in the Neotropics, such as Brazilian agricultural landscapes (Moreira et al., 2015; Saturni et al., 2016; Boscolo et al., 2017; Medeiros et al., 2018; Hipólito et al., 2018; Aristizabal and Metzger, 2019). Furthermore, most studies related to the effects of landscape simplification on biodiversity have focused on alpha diversity, whereas beta diversity has received much less attention (Mori et al., 2018).

According to Baselga (2010) beta diversity quantifies the differences among biological communities and reflects two different phenomena: spatial species turnover and nestedness, which represent the replacement and loss of species between communities, respectively. Beta diversity is an essential approach to elucidate processes involved with changes in community composition due to natural and anthropogenic disturbances such as biological invasions (Socolar et al., 2016; Silva and Hernández, 2018) and agricultural intensification (Gabriel et al., 2006; Karp et al., 2012). Local-field scale studies (alpha diversity) identify only a subset of diversity, whereas beta diversity is a useful tool to quantify all components of diversity at multiple spatial scales (Gabriel et al., 2006). Integrating alpha and beta diversity could accommodate multiple ecosystem services at the landscape level (Frei et al., 2018; Rodríguez-Loinaz et al., 2014) and can aid decision makers and conservationists in selecting appropriate indicators and spatial scales for species conservation (Clough et al., 2007).

We aimed to understand how compositional and configurational landscape heterogeneity influences the diversity of beneficial insects in Brazilian coffee farms. Specifically, we tested whether alpha and beta diversity of beneficial insects change with forest cover and landscape diversity (compositional heterogeneity) and edge density (configurational heterogeneity). We used wasps (Insecta: Hymenoptera: Vespidae), bees (Insecta: Hymenoptera: Apoidea) and flower flies (Insecta: Diptera: Syrphidae) as a model of different groups of beneficial insects. Bees, wasps and flower flies provide important pollination services in natural and agroecosystems (Allen-Wardell et al., 1998; Potts et al., 2016; Inouye et al., 2015; Ollerton, 2017; Lucas et al., 2017, Lucas et al., 2018) including coffee plantations (Roubik, 2002; Ricketts et al., 2004; Klein et al., 2003, Klein et al., 2008; Vergara and Badano, 2009; Saturni et al., 2016; Hipólito et al., 2018). Moreover, bees, wasps and flower flies have been used as bioindicators to assess the loss of biodiversity and the efficiency of restoration and conservation policies (Sommaggio, 1999; Tscharntke et al., 2005; Ricarte et al., 2011; Sommaggio and Burgio, 2014), and many wasp and flower fly species are important agents of biological control of pests in several agroecosystems (Richter, 2000; Rojo et al., 2003; Schmidt et al., 2004; Nelson et al., 2012; Eckberg et al., 2015).

We expected that forest cover, landscape diversity and edge density regulate community composition of beneficial insects such that communities located in coffee monocultural landscapes support subsets of species-rich communities in more heterogeneous landscapes. We also expected that low levels of compositional and configurational heterogeneity result in community homogenization with a few crop-associated species replacing most species. Atlantic Forest is the dominant non-crop habitat in the study region and provides undisturbed nesting habitats for wasps (Souza et al., 2010, Souza et al., 2014, Souza et al., 2015) and bees (Samejima et al., 2004; Siqueira et al., 2012), and specific larval micro habitats for several flower fly species (Medeiros et al., 2018). Bees, wasps and flower flies can be found in both crop and non-crop habitats; consequently, high landscape diversity and edge density may improve the capacity of landscapes to provide multiple resources for beneficial insects via landscape complementation and supplementation. We provide novel information on the effects of landscape composition and configuration on the diversity of pollinator and natural enemy insects in Brazilian farmland.