Responding effectively to accidental introductions of plant pests (e.g., fungi, bacteria, viruses, animals, plants) is complicated because timely and costly decisions must be made across social and ecological scales with limited information. In this dissertation, I provide an interdisciplinary framework that allows responsible institutions to respond quickly and effectively to an emerging, introduced, multi-host pest-pathogen complex using even minimal knowledge available about pest attributes. First, I take an evolutionary ecology approach and examine how the phylogenetic structure of host ranges of different pest-pathogen combinations can be used to predict likelihoods of establishment, spread, and impacts of Fusarium dieback - invasive shot hole borers (FD–ISHB) in the urban-wildland forests and avocado growing regions of Southern California, where the pest-pathogen complex has established after its introduction from Southeast Asia. Phylogenetic dispersion analysis on a comprehensive FD–ISHB host-range data set shows that the strength of the phylogenetic signal is progressively more pronounced for more severely affected host species. As a basis for risk analysis, this understanding helps plant health first responders assess how any polyphagous pest complex might behave when introduced to novel environments with a new set of possible hosts, which in turn informs more efficient and cost-effective phytosanitary surveillance priorities. Second, I conduct a multivariate analysis of fungi and bacteria cultured from wood in a phylogenetically diverse set of live tree hosts to determine if the structure and composition of tree microbiomes is predictive of the likelihood or outcome of attack by FD–ISHB. I further explore interactions within the microbiome between endophytic microbes and the pathogen to identify potential opportunities and mechanisms to shape disease establishment and spread, and evaluate whether endogenous microbes could be manipulated for sustainable integrated pest management. I found consistent differences in wood-inhabiting microbial communities between avocado, which grows in an agricultural setting, and three wildland tree species (willow, sycamore, and oak), but there were no strong, consistent differences among microbial communities based on host attack status. However, enough differences were detected to suggest that inconsistencies most likely reflect undersampling in the community – a common problem with culture-based studies – which sets the stage for future culture-independent studies that integrate a richer data set into the analysis. Furthermore, 15 fungal species and 11 species of bacteria exhibited clear in vitro antagonism against the pathogen, indicating their potential to confer a protective benefit to tree hosts as biological control agents. Finally, I analyze participant-observation and public-document data to assess the effectiveness of governance processes that influence management decisions in a statewide deliberative and consensus-directed process to control FD–ISHB spread and impacts. I found that the comprehensive set of collaborative actions that emerged from this process were due to conditions identified in theoretical frameworks for collaborative governance and could not have been attained by any organization acting alone. These actions were enhanced by the structure and quality of principled-engagement process elements, which benefited from prior histories of cooperation and conflict. Collectively, this dissertation provides valuable technical and collaborative tools to improve integrated pest management and respond to the large-scale socio-ecological disturbances that accompany invasive, introduced pests.