LOOKING AT CITIES THROUGH NOVEL LENSES: STUDYING URBAN BIODIVERSITY ACROSS SPATIAL AND ECOLOGICAL SCALES

Abstract

Land-use change and intensification are main drivers of global biodiversity decline. Urban areas are rapidly expanding worldwide, representing a major threat for biodiversity. Nonetheless, cities are increasingly recognised as ecosystems on their own, harbouring often a considerably high biodiversity, being a major force of eco-evolutionary change, and in turn, providing a wide array of ecosystem functions and contributions to people. Despite an already long history of research, there are still important questions that remain elusive in urban ecology particularly concerning the effects of socio-ecological drivers on urban biodiversity patterns and ecological processes, for example diet preferences, species distributions and interspecific competition. Furthermore, although invertebrate species represent the majority of taxa inhabiting cities, they have been much less investigated than plants and birds, and thus, the diversities patterns for many groups in urban ecosystems are poorly understood. This thesis aimed at exploring the influence of different socio-ecological drivers on different aspects of urban biodiversity (e.g. diversity patterns, species distribution, trait variation, and trophic and competitive interactions) across spatial scales and ecological levels of organisation. To do so, different methodological tools, including macroecological analyses, next-generation sequencing techniques and individual trait-based approaches, have been used to better elucidate the effects of urban socio-ecological drivers on ecological processes and urban biodiversity patterns. Each of the three chapters of this thesis is set at a different spatial scale and ecological level of organisation. Chapter I is part of the European project BioVEINS, and includes 80 study sites in five European cities (i.e. Antwerp, Belgium; Paris, France; Poznan, Poland; Tartu, Estonia; Zurich, Switzerland) distributed following gradients of urban intensity and available habitat, where solitary wild bees and their larval pollen were sampled. Chapter II is part of the project ZuriDiverCity and represents a compilation of four urban ecology studies conducted in the city of Zurich, with 252 study sites representing the most common types of urban greens areas (e.g. parks, brownfields, green roofs, gardens) distributed along multiple urban intensity gradients and data of 1446 animal species from 12 animal groups. Finally, Chapter III is part of the project City4Bees and includes data on managed and wild bees in 23 domestic gardens varying in local habitat features and in the position within the urban intensity gradient in the city of Zurich. Chapter I studies the larval diet preferences of four widespread solitary bee species (i.e. Chelostoma florisomne, Osmia bicornis, Osmia cornuta and Hylaeus communis) using pollen metabarcoding, and, in addition, investigates how larval feeding specialisation affects the distribution of the four wild bee species within urban intensity gradients. It shows that there are different feeding specialisation strategies that can be successful in urban ecosystems allowing bees to exploit a variable part of the urban floral resources. These feeding specialisation strategies are not limited to broad generalism, but also included intermediate levels of generalism, with a certain degree of diet conservatism at the plant family or genus level (in O. bicornis and O. cornuta), and strict specialisation to a single plant genus with few species but widely common in urban habitats (in C. florisomne). However, at the same time, larval feeding specialisation influences the occupation of habitat patches and distribution within urban intensity gradients, and broader diets result in less sensitivity to urban intensity. Broad generalist H. communis was the least sensitive bee to urban intensity. Interestingly, the analyses of its larval diet reveals plant host shifts from a more herbaceous-based diet to a more tree-based diet triggered by losses in the amount of available herbaceous habitat, showing the important role of street trees in providing alternative resources for some wild bee species. Therefore, identifying larval floral preferences could be helpful for identifying key plant taxa and traits for bee survival and for improving strategies to develop bee-friendly cities. Chapter II investigates the ecological properties of urban biodiversity in the city of Zurich using machine learning predictive models and occurrence and abundance data from 1446 species belonging to 12 different animal groups. The occurrence and abundance distribution patterns of the taxonomic groups resemebled those from non-urban ecosystems. Specifically, urban diversity is composed by a very reduced group of hyperdominant and hyperwidespread species in the city while the vast majority of species are scarce and locally- occurring. Furthermore, the application of predictive models revealed that only species with intermediate abundances and occurrences were spatially structured along urban gradients, with decreasing numbers with increasing urban intensity. Strikingly, rare species were less sensitive to urban gradients and their numbers kept constant even in the most utbanised areas of the city. Overall, urban green areas with lower management regimes tended to contain higher species richness of most of the studied groups. The use of predictive models in urban ecosystems, specifically when producing continuous citywide biodiversity maps, is a promising tool to address both ecological questions and inform urban planning and management. Chapter III examines the interplay of resource availability and beekeeping intensity at local and landscape scales in shaping the diversity of urban wild bees in urban gardens in the city of Zurich. By collecting six functional traits and incorporating individual trait variability, the feeding niche partitioning between the wild bee community and the honeybee population was calculated at each site. Lower resource availability at the local scale negatively affected wild bee species richness, whereas lower resource availability at the landscape scale increased honeybee densities at the study sites. In turn, feeding niche partitioning decreased in sites with lower resource availability at the landscape scale (and thus, with larger honeybee densities) and decreased as well with lower resource availability at the local scale, due to a reduction in the number of wild bee species, in particular those that where more functionally similar to honeybees. Conversely, beekeeping intensity at local and landscape scales did not have a direct effect neither on feeding niche partitioning nor on wild bee species richness. Hence, monitoring and management of resource availability is critical to support wild bee diversity and manage urban beekeeping. This thesis provides novel insights on the effects of socio-ecological drivers on selected biodiversity patterns and ecological processes at different scales and ecological level of organisation, through the application of new methodological tools. The application of new analytical tools, such macroecological models or next-generation sequencing technique can pose a great opportunity to revist classical urban ecology questions but with novel lenses. Furthermore, the different results compiled provide a basis to ecologically inform management and planning of urban habitats, specifically to minimize the effects of anthropogenic activities and secure the natural capital of urban ecosystems.