Introduction

The current decline of insect abundance and diversity alerts ecologists and the broad public worldwide (Hallmann et al. 2017, Sánchez-Bayo and Wyckhuys 2019, Wagner et al. 2021). In particular, the loss of pollinating insects has the potential to endanger the entire ecosystem functioning at several trophic levels across ecosystems. Approximately 87% of all wild flowering plants depend on animal pollination (Ollerton et al. 2011), therefore insect pollinators are essential for the preservation of plant biodiversity (Fontaine et al. 2005, Biesmeijer et al. 2006) and present an extraordinarily important economic factor worldwide (Gallai et al. 2009).
The intensification of current agricultural practices is considered to be one of the main driver for the loss of pollinator biodiversity and abundances (Sánchez-Bayo and Wyckhuys 2019, Wagner et al. 2021). The response of pollinators to land-use intensification should differ between pollinator guilds, since taxa highly differ in their ecological requirements and functional traits. Wild bees and hoverflies belong to the main pollinator guilds in agricultural landscapes across different habitats (Stanley and Stout 2013, Rader et al. 2020). Wild bees are often considered to be habitat specialists due to their particular nesting site requirements and their stationary foraging behaviour in addition to their specialized resource uptake of nectar (Westrich 1996, Johnson and Steiner 2010). Hoverflies, in contrast, are less specialized in nectar uptake (Van Rijn and Wackers 2016), foraging across a wide range of habitats and on much larger scales compared to wild bees (Bankowska 1980, Power et al. 2016, Klaus et al. 2021). As a result, hoverflies are regarded as generalists, which are less susceptible to land-use intensification than wild bees (Jaucker et al. 2009, Blaauw and Isaacs 2014, Aguirre-Gutierrez et al. 2015). However, solid empirical evidence is missing (e.g. Jauker et al. 2019) and a recent long-term study reported a catastrophic decline of generalist hoverflies during the past years in Central Europe (Hallmann et al. 2020). Despite recent attempts, our understanding of how wild bees and hoverflies are affected by different measures of land-use intensification is limited, which hampers guidance for conservation measures and forecasting consequences of pollinator losses (Senapathi et al. 2017, Rader et al. 2020).
Land use intensification leads to a higher coverage of arable fields (Maskell et al. 2019). The current management regimes of these arable fields include a high frequency of mechanical disturbance, the application of pesticides and fertilizers. The resulting landscapes barely offer value for pollinating insects as food resources or nesting sites with the exception of short-flowering mass events (Riedinger et al. 2014). As a result, pollinators are restricted to patches of (semi-)natural habitats within the agricultural matrix. Therefore, increasing amount of arable field coverage incorporates a reduction of food supply and habitat loss, which hampers dispersal and (re-)colonization of habitat patches. Consequently, this leads to a decrease of pollinating insects like wild bees (Senapathi et al. 2017). In contrast, some studies reported positive effects of arable field cover on hoverflies in agriculture landscapes (Haenke et al. 2009, Gabriel et al. 2010, Brandt et al. 2017). Though these mechanisms are not fully understood, it can be expected that wild bees negatively and hoverflies positively respond to arable field cover.
Moreover, land use intensification may cause a loss of landscape heterogeneity (Maskell et al. 2019). The reduction of habitat diversity at the landscape scale reduces the number of potential niches and food resources, thus, landscape homogenization decreases species diversity (Fahrig et al. 2011, Senapathi et al. 2017). Although landscape heterogeneity and arable field cover may often negatively related to each other (Tscharntke et al. 2012), high landscape heterogeneity may compensate negative effects of arable field cover (Maskell et al. 2019). However, it remains unclear how the effect of both parameters changes with spatial scale and which is of greater importance for both pollinator guilds (but see Maskell et al. 2019). Hoverflies may suffer more from landscape homogenization, as they disperse across a wider range of habitats compared to wild bees that forage nearby their nests.
Land-use intensification may reduce the habitat quality of pollinators. Direct and indirect soil fertilization decreases overall plant species diversity (Maskell et al. 2010, Borer et al. 2014), often accompanied with a particular loss of forbs in grasslands (Maskell et al. 2010). This decline in plant diversity is also found in the context of land abandonment of unproductive habitats, such as dry grasslands, as a consequence of land-use intensification and the (subsequent) cessation of traditional land use practices (Habel et al. 2013). The decline of plant diversity may have a negative effect on pollinator diversity, since many pollinator species show a strong specialization towards particular flower traits (Fenster et al. 2004, Fontaine et al. 2006, Fornoff et al. 2017). However, rather than the taxonomic diversity of plants per se , the functional diversity of flower traits should positively affect pollinator diversity (Fontaine et al. 2006, Fornoff et al. 2017). Moreover, particular flower traits that attract pollinators in the landscape may increase the local pollinators. So far, detailed analyses of flowering traits on pollinators are missing in the landscape context, which is an essential part of how land-use intensification affects local habitat quality for pollinators. Hereby, functional flower diversity should have a stronger effect on wild bees, because they show a stronger specialization to specific flower traits compared to hoverflies (Johnson and Steiner 2000, Van Rijn and Wackers 2016). Otherwise flower traits related to attractiveness should have a stronger effect on hoverflies that migrate through the landscape.
In this study, we aim to reveal responses of two important pollinator groups to different measures of land use intensification, in order to get a better understanding of the underlying mechanisms of the current pollinator loss and subsequent ecosystem functioning. As a study system, we used isolated dry grassland patches that are embedded in an otherwise intensively used agricultural landscape in NE Germany. We sampled bees and hoverflies at 22 dry grassland patches within three sampling campaigns using pan traps. Further, we quantified the local flowering plant community at the time of sampling and estimated different measures of local flower diversity and ‘attractiveness’. We determined arable field cover and landscape heterogeneity on consecutive radii from 60m – 3000m around the dry grassland sites, in order to reveal the ‘scale of effect’ (Jackson and Fahrig 2015), i.e. the spatial scale at which the predictor has the strongest influence on the response variable.
We hypothesize that
1) the proportion of arable field cover surrounding the dry grassland sites has a negative effect on wild bees (species richness and abundance) and a positive effect on hoverflies,
2) landscape heterogeneity has a stronger positive effect on hoverflies compared to wild bees,
3) the spatial scale at which arable field cover and landscape heterogeneity affect the pollinator guilds, is smaller for wild bees than for hoverflies,
4) functional flower diversity positively affects wild bees in particular and flower traits that are associated with ‘attractiveness’ positively affect hoverflies in particular.