1. Fragmentation per se due to human land conversion is a landscape-scale phenomenon. Accordingly, assessment of distributional patterns across a suite of potentially connected communities (i.e. metacommunity structure) is an appropriate approach for understanding the effects of landscape modification, and complements the plethora of fragmentation studies that have focused on local community structure. To date, metacommunity structure within human-modified landscapes has been assessed with regard to nestedness along species richness gradients. This is problematic because there is little support that species richness gradients are associated with the factors moulding species distributions. More importantly, many alternative patterns are possible, and different patterns may manifest during different seasons and for different guilds because of variation in resource availability and resource requirements of taxa. 2. We determined the best-fit metacommunity structure of a phyllostomid bat assemblage, frugivore ensemble, and gleaning animalivore ensemble within a human-modified landscape in the Caribbean lowlands of Costa Rica during the dry and wet seasons to elucidate important structuring mechanisms. Furthermore, we identified the landscape characteristics associated with the latent gradient underlying metacommunity structure. 3. We discriminated among multiple metacommunity structures by assessing coherence, range turnover, and boundary clumping of an ordinated site-by-species matrix. We identified the landscape characteristics associated with the latent gradient underlying metacommunity structure via hierarchical partitioning. 4. Metacommunity structure was never nested nor structured along a richness gradient. 5. The phyllostomid assemblage and frugivore ensemble exhibited Gleasonian structure (range turnover along a common gradient) during the dry season and Clementsian structure (range turnover and shared boundaries along a common gradient) during the wet season. Distance between forest patches and forest edge density structured the phyllostomid metacommunity during the dry and wet seasons, respectively. Proportion of pasture and forest patch density structured the frugivore metacommunity during the dry season. 6. Gleaning animalivores exhibited chequerboard structure (mutually exclusive species-pairs) during the dry season and random structure during the wet season. 7. Metacommunity structure was likely mediated by differential resource use or interspecific relationships. Furthermore, the interaction between landscape characteristics and seasonal variation in resources resulted in season-specific and guild-specific distributional patterns.