Ecological phenomena act on various spatial and temporal scales. To understand what causes animal populations to build and decline depends heavily on abiotic and biotic conditions which vary spatiotemporally throughout the biosphere. One excel- lent example of animal populations dynamics is with locusts. Locusts are a subset of grasshoppers that undergo periodical upsurges called swarms. Locust swarms have plagued human history by posing significant threats to global food security. For example, the 2003-2005 desert locust (Schistocerca gregaria) swarm destroyed 80%-100% of crops in the impacted areas and cost over US $500 million in mitigation as estimated by the Food and Agriculture Organization of the United Nations. An integrative multi-scale approach must be taken to effectively predict and manage locust swarms. For my dissertation, I looked at the ecological causes of locust swarms on multiple scales using both the Australian plague locust (Chortoicetes terminifera) and desert locust as focal species. At the microhabitat scale, I demonstrated how shifts in the nutritional landscape can influence locust gregarization. At the field level, I show that locust populations avoid woody vegetation likely due to the interactive effect of plant nutrients, temperature, and predators. At the landscape level, I show that adaptations to available nutrient variation depends on life history strategies, such as migratory capabilities. A strong metapopulation structure may aid in the persistence of locust species at larger spatial scales. Lastly, at the continental scale I show the relationship between preceding vegetation and locust outbreaks vary considerably between regions and seasons. However, regardless of this variation, the spatiotemporal structure of geographic zone > bioregion > season holds constant in two locust species. Understanding the biologically relevant spatial and temporal scales from individual gregarization (e.g. micro-habitat) to massive swarms (e.g. landscape to continental) is important to accurately predicting where and when outbreaks will happen. Overall, my research highlights that understanding animal population dynamics requires a multi-scale and trans-disciplinary approach. Into the future, integrating locust re- search from organismal to landscape levels can aid in forecasting where and when locust outbreaks occur.

The aboveground surfaces of plants (i.e. the phyllosphere) comprise the largest biological interface on Earth (over 108 km2). The phyllosphere is a diverse microbial environment where bacterial inhabitants have been shown to sequester and degrade airborne pollutants (i.e. phylloremediation). However, phyllosphere dynamics are not well understood in urban environments, and this environment has never been studied in the City of Phoenix, which maintains roughly 92,000 city trees. The phyllosphere will grow if the City of Phoenix is able to achieve its goal of 25% canopy coverage by 2030, but this begs the question: How and where should the urban canopy expand? I addressed this question from a phyllosphere perspective by sampling city trees of two species, Ulmus parvifolia (Chinese Elm) and Dalbergia sissoo (Indian Rosewood) in parks and on roadsides. I identified characteristics of the bacterial community structure and interpreted the ecosystem service potential of trees in these two settings. I used culture-independent methods to compare the abundance of each unique bacterial lineage (i.e. ontological taxonomic units or OTUs) on the leaves of park trees versus on roadside tree leaves. I found numerous bacteria (81 OTUs) that were significantly more abundant on park trees than on roadside trees. Many of these OTUs are ubiquitous to bacterial phyllosphere communities, are known to promote the health of the host tree, or have been shown to degrade airborne pollutants. Roadside trees had fewer bacteria (10 OTUs) that were significantly more abundant when compared to park trees, but several have been linked to the remediation of petroleum combustion by-products. These findings, that were not available prior to this study, may inform the City of Phoenix as it is designing its future urban forests.