fish community composition in freshwater lakes is shaped by a range of biotic and abiotic factors, including environmental conditions, species interactions, and spatial connectivity between waterbodies. while aquatic community ecology studies historically treated lakes as isolated systems, recent research has increasingly embraced a metacommunity framework, integrating spatial connectivity with environmental and biological predictors of community composition. despite this shift, few studies have thoroughly examined the relative roles of spatial connectivity, environmental factors, and species interactions in shaping lake fish communities. to address this gap, i conducted a study across 81 lakes distributed within two quaternary watersheds at the iisd experimental lakes area in northwest ontario. using joint species distribution modeling (jsdm) alongside spatial eigenvector mapping techniques—asymmetric eigenvector mapping (aem) and moran’s eigenvector mapping (mem)—drivers of fish community composition were investigated. results indicate that spatial variables—specifically lake connectivity, stream flow direction, and the maximum gradient along connecting streams—are primary drivers of fish metacommunity composition. in presence-absence models, these spatial factors explained more variation than environmental variables and species co-occurrence patterns (potentially reflecting species interactions). conversely, relative abundance models (conditional on presence) performed poorly across all ecological models evaluated. these findings provide valuable insights into the role of spatial connectivity relative to other factors in shaping fish community structure on a presence-absence basis, emphasizing the importance of applying a metacommunity approach in community analyses.