Open Collections

Abstract

Urban expansion is rapidly and drastically altering the environment. To mitigate biodiversity loss and to maintain ecosystem services where we live, it is critical that we protect biodiversity in growing urban areas. Supporting insect pollinators in urban landscapes is especially important, as these critical ecosystem players are facing widespread declines. In my thesis, I first investigated how urban landscape composition impacts pollinator biodiversity. To model associations between landscape and biodiversity, I applied occupancy models that integrated community science and natural history collections data while accounting for detection biases. Across a continental extent, I discovered that pollinator diversity was positively associated with the amount of natural greenspace in an urban landscape, emphasizing that expanding cities should prioritize preservation of forest/grassland remnants and greenbelt lands. Accounting for greenspace area, the proportion of racial minorities in an urban landscape was negatively associated with pollinator diversity, consistent with the hypothesis that legacies of unjust socioeconomic policies in neighborhoods with larger racial minority populations compromise biodiversity outcomes. Future urban management improvements should prioritize action in these neighborhoods to foster equal access to biodiversity and ecosystem services. Next, I examined how pollinator community dynamics respond to urban park restorations, specifically the transformation of turfgrass lawns into flower meadows. Applying temporally explicit occupancy models to field data, I found that restored parks increase and sustain species richness over the long run. I also found that existing flowering woody plants increased colonization by specialist pollinators species, many of which are of high conservation concern. Finally, I conducted a study to examine the impacts of habitat on pollinator detectability. A mark-recapture approach demonstrated that restored habitats with more flowers and taller vegetation increased pollinator detection rates. Consequently, conventional generalized linear models that did not account for imperfect detection overestimated the impacts of habitat improvements on pollinator abundance. I showed that increasing sampling intensity rather than adding more study sites can increase the accuracy of conventional methods for estimating hypothesized drivers of abundance. Together, my thesis provides guidance on how to better design, manage and improve ecological studies in urban landscapes to promote pollinators and the benefits that they provide.