DISCUSSION
Pollination is an important ecosystem function for both natural and anthropogenic habitats yet, the effects of urbanisation on pollinators and pollination are poorly known. In this study, we reviewed the literature and performed a meta-analysis on the effects of urbanisation on pollinators and pollination. Our meta-analysis revealed an overall negative effect of urbanisation on pollinator abundance and richness. The magnitude of the effect was dependent on the taxonomic group of the pollinator, its origin and functional traits. Furthermore, flowering plant richness had a positive effect on pollinator richness revealing the importance of local habitat resource availability for pollinators. Our meta-analysis also revealed that pollination services are enhanced in urban areas most likely due to the high abundance of generalist and managed pollinator species. Below, we expand on these results and discuss their implications for pollinator conservation in cities.
Urbanisation results in a drastic modification of habitats with negative effects on biodiversity (McKinney 2008). The negative effects of urbanisation on both pollinator abundance and richness, reported in our study, are in line with previous meta-analyses (Fenoglioet al. 2020; Millard et al. 2021). The increase in impervious surfaces, habitat loss, fragmentation, and degradation as well as environmental pollution are the primary hypothesised drivers for the observed reduction of pollinator biodiversity in cities (González-Varo et al. 2013; Vanbergen & Initiative 2013). In addition to the overall negative effects on pollinator biodiversity, our results revealed that pollinator taxonomic groups differ in their sensitivity to urbanisation. Lepidoptera was the taxonomic group that was found to be most affected by urban development. Many butterflies and months require specific host plants for larval development and nectar consumption as adults and they appear to be very sensitive to urban environmental stressors such as heat island effects, air and light pollution (Ramírez-Restrepo & MacGregor-Fors 2017; Fenoglio et al. 2020; Callaghan et al. 2021). Due to their sensitivity to human disturbances, our results further highlight the suitability of Lepidoptera as a bioindicator taxon to quantify responses to urbanisation. Our results further demonstrate that Lepidoptera is the taxonomic group most at risk from urbanisation and should be prioritised for conservation in cities.
The climatic region of the study was not an important moderator of the effects of urbanisation on pollinators. As has been shown in a previous study (Fenoglio et al. 2020), urbanisation had a negative effect on pollinator richness in both tropical and non-tropical cities. Urban development is a global phenomenon that leads to convergence of urban environments (Santangelo et al. 2022). This global convergence of city environments could be the main driver of the observed declines of pollinator species richness irrespective of the climatic region of the study. Nonetheless, urbanisation significantly affected the abundance of pollinators only in non-tropical regions. This lack of an effect could be due to the relatively small number of studies conducted in the tropics. Alternatively, it might reflect the high levels of urbanisation and thus higher impact on pollinators in non-tropical temperate regions (Faeth et al. 2011; Elmqvist et al. 2013).
Species responses to urbanisation are trait-specific (Buchholz and Egerer, 2020; Wenzel et al. 2020), and the implementation of ecological trait approaches for urban biodiversity conservation provides a mechanistic understanding of the relationship between biodiversity and urban environmental constrains. Urban landscapes can act as environmental filters for pollinator species depending on their ecological traits, facilitating or hindering their colonisation and survival in cities (Buchholz & Egerer 2020). In our meta-analysis, we found that the pollinator traits mainly affected by urbanisation are those related to nesting behaviour and sociality. In regard to nesting behaviour, our analysis confirmed the hypothesis that below ground nesting bees are negatively affected by urbanisation (Neameet al. 2013; Geslin et al. 2016). The increase in impervious surfaces with urban development and the intensive management of urban green land-uses result in a reduction in the availability of suitable habitat (i.e., bare soil) for ground nesting bees (e.g., Andrenidae, Halictidae, Colletidae; Potts et al. 2005; Pereiraet al. 2021). Sociality is also hypothesised to be an important trait related to urban environmental filters (Wenzel et al.2020). In our meta-analysis, we found the abundance of solitary bees to be negatively affected by urbanisation. The lack of ecological and behavioural flexibility as well as their relatively small population sizes compared to social bees might be the main drivers of solitary bee abundance declines with urbanisation (Chapman & Bourke 2001; Banaszak-Cibicka & Zmihorski 2012). Although the pollinator activity length was not affected by urbanisation, we found seasonal variation in the effects of urbanisation on pollinator communities. Urbanisation had strong negative effects on the abundance and richness of spring pollinator communities. As suggested by previous studies, this might be driven by the scarcity of spring-blooming shrubs and trees and the overall lack of spring-foraging resources for pollinators in cities (Matteson et al. 2008; Banaszak-Cibicka & Zmihorski 2012). Solitary bee species that are ground-nesters and spring flyers (e.g., Andrenidae) are at increased risk due to urbanisation, and conservation efforts should be primarily directed towards increasing spring floral resource availability as well as nesting opportunities for those pollinators in cities.
Pollinators of native or non-native origin also responded differently to urbanisation. Non-native pollinator species richness increased, and native pollinator species richness decreased with urbanisation. Our results suggest that cities might be hotspots of non-native pollinator species (Normandin et al. 2017; Fitch et al. 2019). This phenomenon could be driven by species that are introduced into cities through human activities, for example, urban beekeeping (Egerer & Kowarik 2020). Following an introduction, a species’ diet breadth, nesting behaviour and thermal tolerance might further facilitate its establishment in cities. Generalist species with strong preferences for exotic flowering plants, cavity nesters as well as species with thermal tolerance that matches the urban conditions are usually good urban invaders (Goulson 2003; Hamblin et al. 2017). However, it is worth noticing that non-native pollinators may exacerbate conservation risks to native wild species by competition for floral resources, nest sites, or transmission of parasites and pathogens (Fitch et al.2019).
Pollinators differ in their life-history traits however they all depend on floral food resources for their survival (Willmer 2011). Floral resources are a limiting factor for the populations of all pollinators. Furthermore, there is a strong relationship between flowering plant species richness and pollinator richness in a community (Ollerton 2017). Our meta-regression analyses revealed, this trivial, although rarely explored intrinsic link between flowering plant richness and pollinator diversity. This relationship is of great conservation importance as it reinforces current pollinator initiatives that argue for flower plantings to promote pollinator diversity.
In our meta-analysis, pollination services, estimated either as seed set or flower visit duration, increased with urbanisation. Pollination could be influenced by pollinator visit quantity and “quality” as well as by the structure of the local flowering plant community that could influence visitation rates (Bruckman & Campbell 2014). The increase in flower visit duration as well as seed set in cities point to the importance of the “quality” of the pollinator in terms of the number of compatible pollen grains deposited on stigmas (Ne’eman et al. 2010). The increase in flower visit duration in cities suggests a lack of a dilution effect and pollination efficiency due to the potentially high supply of floral resources in urban green land uses. Plants with radial flower morphology particularly benefited from urbanisation. Plants with radial flowers and a central cluster of anthers typically have shallow, exposed nectaries, making both nectar and pollen easily available to flower visitors (Willmer 2011). Plants with these floral traits are typically thought of as generalists (Ollerton et al. 2007) and are visited by many wild pollinator taxa including generalist and managed bee species that could be abundant in cities.
The positive effect of urbanisation on pollination was driven by generalists (bumble bees) and managed bee species (honey bees). Honey bees and bumble bees are generalist pollinators and the most important pollinators in both natural and agricultural ecosystems (Garibaldiet al. 2013; Kleijn et al. 2015; Hung et al. 2018). The increase in honey bee visitation rates in cities is probably due to urban beekeeping. Urban beekeeping is currently booming with both benefits and negative implications (Ropars et al. 2019; Sponsler & Bratman 2021). Urban beekeeping might help with the pollination of urban agricultural crops and wild flowering plants and provide locally produced honey, however, it might also lead to resource competition with wild pollinators and to increased transmission of diseases between honey bees and non-Apis bees (Ropars et al. 2019; Proesmanset al. 2021; Sponsler & Bratman 2021). We argue that city authorities should regulate the intensity of urban beekeeping and future studies should further investigate the benefits and negative aspects of urban beekeeping for wild pollinators and pollination in cities. Nonetheless, based on our results, it appears that urban generalists and managed pollinators provide adequate pollination services to flowering plants.
In conclusion, our results provide evidence that urbanisation has negative effects on pollinator communities however these are not necessarily translated into negative effects on pollination service provision. While we acknowledge the small sample size for some of our moderators, our meta-analysis further highlighted that Lepidoptera are the pollinator group most affected by urbanisation, and that pollinator ecological traits as well as local floral resource richness could mediate the responses of pollinators to urbanisation. Regarding pollination services, our results point towards the importance of generalist bees as well as the managed honey bee for plant reproduction in cities. Overall, more research in cities is needed to guide conservation actions and policies for pollinators and pollination for food security and wild flower reproduction.