The importance of spatial aggregation and density dependence have been extensively explored in ecology. However, both factors might interact or modify one another to differentially influence net population or community-level dynamics. In a desert annual plant community, we experimentally tested for these effects at nested spatial scales by planting different number of patches within a plot but holding plot-level seed density constant, by manipulating the number of occupied patches adjacent to a target patch, and by varying local seed density within patches. In doing so, we addressed the following specific questions: (i) is there an effect of the number of patches of seeds planted on plot-level community measures, (ii) are there patch–patch interactions within plots, (iii) are there density-dependent responses to changing initial seed density within patches, and (iv) do patterns observed at finer spatial scales predict patterns at larger scales? Community parameters measured at the plot level-emergence, plant density, and mean plant size-were negatively affected by increasing seed aggregation into fewer patches within a plot. However, converse effects were detected at finer spatial scales. The presence of neighbouring patches increased emergence within a plot, while increased seed density within patches decreased emergence but strongly increased survival and mean plant size. Hence, the net negative effects detected at the arbitrary plot level do not necessarily correspond to the interactions occurring at finer spatial scales within the plot. These findings suggest that the use of single scales of inquiry can potentially misrepresent the importance of a particular interaction, such as competition, and that the experimental consideration of finer scales within a community reveal different patterns and do not predict final net effects within a community.