Causal effects of roads on community assembly in tropical forests

Several studies have highlighted the impacts of roads and highways on species dispersal, abundance, and persistence, since these are the infrastructures that most threaten biodiversity. Yet, little is known about how roads affect community assembly processes at broader spatial extents. Understanding interactions between roads and community assembly processes is an urgent need especially in tropical areas, where key biodiversity hotspots and ecosystem services are critically threatened. Here, we review how roads could shape community assembly processes and (meta)community structure across biogeographic extents for animals, with a focus on identifying causal effects. We use causal diagrams to illustrate the expected relationships among variables representing processes influencing species composition and dissimilarity under road influence. In doing so, we highlight how observational studies on road effects are prone to confounding and bias, using the concrete example of the controversial BR-319 highway in the Amazon. We propose the use of quasi-experimental designs as an alternative for better assessing the impacts of roads’ construction on animal communities. We conclude that more integrated and multi-process approaches are needed to study (meta)communities, moving from mere descriptions of correlations to the identification of the causality of ecological processes. This can be more easily achieved by integrating spatial and temporal ecological data.