A large-scale dataset reveals taxonomic and functional specificities of wild bee communities in urban habitats of Western Europe

Here we assessed how species and functional diversity components of wild bee assemblages responded to increasing urbanization levels, using a large dataset encompassing recent surveys gathering 838 sampling sites located in natural, semi-natural and urban habitats of France, Belgium and Switzerland.

We found a weak, but significant negative effect of the proportion of impervious surfaces in a 500 m radius around each site on local species richness of bee communities. Thus, sites with high soil sealing tended to host less species than those with low soil sealing. However, this trend was not observed when using human population density as an urbanization metric: sites with denser human populations hosted on average the same number of species as less densely populated sites.

Concerning taxonomic homogenization of communities, we did not record any effects of urbanization, both in terms of impervious surfaces or human population density.

Analyses of occurrence rates of bee functional traits revealed significant differences between poorly and highly urbanized communities, for both urbanization metrics. With higher human population density, probabilities of occurrence of above-ground nesters, generalist and small species increased, and a higher probability of occurrence of above-ground nesters, generalists and social bees were recorded in areas with high soil sealing.

Therefore, we found overall consistent results linking urbanization and wild bees taxonomic as well as functional trait diversity, even though analyses stemmed from a combination of many independent studies covering a broad range of anthropized and natural aeras from western Europe. This further highlights the greater generalizability of those ecological trends throughout European temperate biomes compared to other studies typically focusing on a single city and its immediate vicinity.

Two complementary metrics of urbanization intensity

To quantify urbanization, we used two variables: soil sealing^12,16,19,36 in a 500 m radius, and the mean human population density, also in a 500 m radius, the latter variable being used only recently to assess pollinator responses to urban environments^37,38. These two variables return different but complementary information concerning urban environments. Indeed, if soil sealing gives an idea as to how human activities impact land use, human population density helps distinguish between very dense urban areas and very impervious areas with lower densities of buildings. High human population density areas are usually associated with high levels of soil sealing, but the contrary is not true. Similarly, areas with low soil sealing are usually associated with low human population densities, but again, the opposite is not always true. Therefore, we found it informative to consider both variables when analyzing the response of wild bee assemblages to urbanization.

Note that some specific habitat types, for example business districts, are exceptions to the rule. These places are indeed very densely urbanized, but with very low population density. However, no inventories have been carried out in these places, and thus will not be a problem for our study.

Response of bee community species richness to urbanization

One of our goals was to position this study in the context of the contrasting findings on pollinator communities and urbanization. Whereas no consistent trend is reported in literature¹⁵, our large dataset reveals that high soil sealing is detrimental to wild bee species richness. This offers a unified view of a trend that has been unequally evidenced from studies focusing on a single or few cities only. High proportions of soil sealing reduce the availability of nesting sites for ground-nesting bee species. This may in turn lower the species diversity of local assemblages, by filtering out ground-nesting bees, leaving mainly cavity-nesting bees. Furthermore, high levels of soil sealing can lead to depletion of floral resources, of extreme importance for bees, especially in highly disturbed environments such as cities^39,40. Note that several previous studies report the opposite, with high local species richness of wild bees in urbanized habitats. However, these positive effects are often associated with intermediate levels of urbanization^15,16, where private gardens and other green spaces may supply abundant floral resources, in conjunction with intermediate levels of soil sealing^{16,17,18,19,20,24}.

On the contrary, there was no significant relationship between local species richness and human population density. Recently, two recent studies have used this metric to analyze how urbanization impacts local diversity of bee, hoverfly³⁷ or butterfly³⁸ assemblages, and both studies report negative impacts of human population density. However, high levels of human population density do not necessarily correlate with low availability of floral resources or nesting sites for pollinating insects. Several studies show that densely-populated urban environments may be adequate habitats for pollinating insects, due to alternative management practices of urban green space⁴¹ and the year-round availability of ornamental flowers^42,43. Here, the absence of a clear effect of human population density on local bee species richness masks a change in the species composition of the communities, as shown by the increasing proportion of cavity nesters, compared with ground nesters. Indeed, despite the lower availability of nesting resources for ground-nesters, cavity-nesters take over in high-density areas, where more concrete structures and buildings are present¹⁵, thus they may compensate for the loss of ground-nesting bee species.

Wild bee community homogenization and urbanization

We did not observe any relationship between mean pairwise β-diversity and the two metrics of urbanization. This result contrasts with those of Banaszak-Cibicka and Żmihorski (2020)⁴⁴ who found more homogeneous wild bee communities in urban environments compared to non-urban ones. Similar results have been reported for bees, with homogenization of urban pollinator communities compared to rural ones^28,45. Biotic homogenization in urban environments has also been reported for other taxa, for example birds⁴⁶.

In our study, when considering urbanization levels, either in terms of soil sealing or human population density, urban wild bee communities are not more or less taxonomically homogeneous than non-urban ones. It is important to note that this result does not imply that urban and non-urban wild bee communities are similar, but that the homogenization of wild bee communities is constant throughout the urbanization gradient. In other words, urban communities are as dissimilar as non-urban ones. Here, the β diversity values are quite high (ranging from 0.68 to 0.96), emphasizing that even urban areas have quite dissimilar communities when compared to each other. This high level of dissimilarity among wild bee communities in urban environments can be explained by the large range of biogeographical regions encompassed in our dataset (Fig. 5), as each of these regions harbors a specific wild bee fauna^{<a data-track="click" data-track-action="reference anchor" data-track-label="link" data-test="citation-ref" aria-label="Reference 34" title="Nieto, A. et al. European red list of bees. (Publications Office, 2014). at ” href=”https://www.nature.com/articles/s41598-022-21512-w#ref-CR34″ id=”ref-link-section-d44723685e2749″>34}.

Local factors in cities might also explain these high levels of dissimilarity. We know for example that green space connectivity has effects on species richness, with more wild bee species and abundance in cities with more connected green spaces⁴⁷. Another local explanation might come from contrasting green space management practices among cities. Not all cities have the same policies, and urban green space management is crucial to the establishment and sustainability of diverse pollinator communities^14,15,48. Thus, we expect more dissimilar wild bee communities among cities with differing green space layout and management.

Functional responses of bee communities to urbanization

Several studies have already shown trends on how urban areas filter wild bee communities based on their functional traits (see³⁰ and⁴⁹ for reviews). However, as for taxonomic diversity, it is often difficult to identify clear variation patterns⁵⁰. Using our large dataset, we could identify typical wild bee functional traits that are favored in urban environments, thus informing on the average functional profiles of wild bee species that may thrive in cities. We found urban wild bees in general to be typically above-ground nesters and generalists, while different trends were established for their body size and sociality, depending on the considered urbanization metric (Fig. 6).

Nesting habits

Above-ground nesting species were more frequent with increasing urbanization than below-ground nesting ones, and this result was recorded with both urbanization metrics.

This result is consistent with what was previously reported in the literature^16,49,51,52. Indeed, cities, with high proportions of impervious surfaces and buildings, offer fewer nesting habitats to ground-nesting species¹⁵, nesting sites becoming a limiting factor³⁹. On the other hand, above-ground nesters can do well in cities with the presence of man-made structures, depending on their ability to use them and on their availability⁵³.

The presence of green areas in cities can help ground-nesting bee species by offering more nesting opportunities and resources¹⁷. Several studies highlight the importance of parks and gardens in supporting bee biodiversity in cities^12,18,31,54, which otherwise are constraining environments due to soil sealing.

Diet

Generalist species were more frequent in more urbanized sites than specialist ones, and this was recorded for both urbanization metrics.

This is in accordance with what was previously found in the literature^{32,50,51,52,54,55}, as specialist bee species depend on the presence of their host plants to complete their life-cycle, which are often scarce due to the rarefaction of native flowering resources. As one can find many exotic flowers in cities, especially in residential gardens and urban parks⁵⁶, we expect to detect less oligolectic bee species in densely urbanized habitats⁵⁷.

Notwithstanding, Banaszak-Cibicka et al. (2018)²⁰ found more oligolectic species in urban parks of Poznań (Poland) compared to a national park. Thus, urban areas are not always depleted of specialist species, and well-managed parks with preserved native floral resources can obviously support specialist wild bee species in cities⁵⁸.

Additionally, it is important to emphasize that the presence of an exotic plant species may concomitantly support an associated specialist bee species. In Poland, for instance, the spread of Bryonia dioica in urban environments also brought the Andrena florea wild bee species, specialized on this plant⁵⁹.

Body size

We recorded contrasting effects of the two urbanization metrics on wild bee body size: small species were more frequent in relation to higher human population density compared to large species, but we found no difference with the proportion of impervious surfaces. Contrasting impacts of urbanization on bee body size are also reported in the literature, with some studies finding little to no effect^32,50, and some finding that urbanization often favors smaller bee species^12,30,60. Bee body size is of particular importance because it is related to the foraging range of individuals^61,62. In fragmented habitats, such as dense urban environments, distances between suitable nesting and feeding habitats may select for smaller species that can remain on small green spaces and rarely need to commute across several green spaces. Furthermore, small bees may be favored given that they need fewer floral resources than large bees, even though large bees can fly further⁶².

This might also explain the difference in the response of bee body size to the two urbanization metric results. In densely populated cities, it is harder to fly between suitable habitats, even for larger bees, as higher buildings and structures may act as barriers to their movement. Indeed, it has been recently shown that the 3D structure of cities impacts wild bee community composition⁶³. Thus, being able to fly further might no longer be an advantage, and larger bees, requiring more floral resources than smaller ones, might be selected against. On the contrary, very impervious areas do not always host high building density (for example, as in the case of parking lots), thus making it easier for large wild bees to fly between bare soil areas.

Densely populated areas might also exhibit warmer temperatures due to the urban heat island effect, and this could, in turn, result in the selection of smaller individuals, as we know that in cities, higher temperature results in smaller body sizes⁶⁴.

Sociality

We also recorded contrasting effects of the two urbanization metrics on sociality: social species were more frequent in relation to higher proportion of impervious surface compared to solitary ones, but no effect was recorded with human population density. This is in agreement with a recent literature review that reports on no consensus concerning the response of this trait to urbanization³⁰.

However, some urban habitats are shown to host more social species than rural habitats^20,32, which may be linked to better reproductive success in cities compared to rural habitats such as agricultural environments⁶⁵, an explanation that is consistent with our results on the soil sealing—sociality relationship.

Conclusion, limits & future directions

Overall, our findings suggest that urban environment filters wild bee communities based on their functional traits. Our results also underscore different impacts of urbanization metrics on local species diversity, with a significant negative impact of soil sealing. On the contrary, both soil sealing and human population densities create strong functional filtering of trait assemblages.

These results are particularly relevant since they arise from a range of independent studies, thus providing a general view on the wild bee communities in urban environments from western Europe. Since this study covers different biogeographical zones, it further underlines its applicability to other temperate countries. We therefore expect similar patterns to shape wild bee communities in urbanized areas from other temperate regions, but further confirmatory studies would be welcome.

Our study also delivers a clear message concerning wild bee communities in urban environments. Urban environments cannot compare with non-urban ones in terms of species richness and trait diversities of bee communities. However, simple management practices of urban green spaces, such as differentiated management, or simply low management⁶⁶, may help in maintaining this diversity. Indeed, not all green spaces are equally valuable in supporting wild bees, and pollinator assemblages in general⁴⁹. For example, it has been shown that pollinator richness was positively influenced by green space size, but also by management measures such as mowing⁶⁷. Increasing the quantity of floral resources and their spatio-temporal availability and diversity^40,68 could also help conserving pollinator communities and pollination function in cities⁶⁹, as long as these resources are native or attractive to pollinators.

We can then hypothesize that changes in managing practices could help increase functional diversity of bees in cities, with specialist and ground-nesting species being found more frequently in these low-managed urban areas.

Finally, if managing urban green space is of great importance to protect biodiversity in cities, it is crucial to involve all stakeholders, especially residents⁷⁰ to achieve efficient and socially-accepted measures.

In the future, it will be important to consider intra-city landscape variation, and see how urban characteristics might influence taxonomic and trait diversity. This will surely allow us to better understand the dynamics shaping wild bee communities in urban environments.

Source: Ecology - nature.com