Ecological complex networks representing predator-prey interactions are crucial for analyzing the structure, function, and dynamics of ecosystems. However, gathering data on what species eat is challenging and resource intensive, and the difficulty of capturing the interactions across diverse taxa and broad spatio-temporal scales has confined existing networks taxonomically or geographically. To address these challenges, we explored an analytical approach to a predator-prey network constructed using crowd-sourced observations encompassing thousands of interactions and taxonomic groups across multiple spatio-temporal dimensions, which we present here. Our work is among the first to model a large-scale global foodweb from crowdsourced data and underscores the potential of such networks to uncover structurally important species across food chains, trophic levels and motifs, despite the extensive gaps in data collection. This work demonstrates the opportunities and challenges of combining citizen science with ecological network analysis while candidly assessing limitations of using crowdsourced data for holistic ecosystem research.