The role of habitat‐forming species in promoting biodiversity is widely acknowledged to vary across environmental gradients according to the extent to which they modify resources and environmental conditions. Population‐ and individual‐level traits of habitat‐forming species that influence species interactions may vary across gradients, but the importance of this indirect effect of environmental context is seldom considered. Here, we conducted surveys and field experiments to partition the effects of wave exposure on habitat‐provisioning for invertebrates by oysters into direct and indirect effects, arising from morphological variation of the oysters. A survey of nine sites with varying degrees of wave exposure in Port Jackson, Australia revealed a decline in oyster densities and surface area as wave energy increased. Correlated to declining oyster surface area was a decrease in the richness and abundance of associated invertebrates. By contrast, taxon diversity increased with increasing wave energy. Experimental deployments of oysters at high and low wave energy sites confirmed that variations in oyster morphology was a phenotypically plastic response to environmental conditions. Oyster recruitment was also lower at high as compared to low wave energy sites, further contributing to the variation in oyster habitat among sites. A colonisation experiment in which exposed and sheltered morphologies of oysters were deployed under high and low wave energy conditions in a fully orthogonal design found that invertebrate communities were influenced by both the wave energy of sites and by habitat structure. Our study suggests that in some instances the indirect effects of environment on habitat availability, arising from changes in habitat‐forming species density and morphology, may be as, or even more, important than the direct effects. Understanding how traits of habitat‐forming species respond to environmental conditions, and how intraspecific trait‐variation cascades to influence associated communities is critical to predicting when and where positive species interactions will be greatest. [ABSTRACT FROM AUTHOR]