Abstract Aim The aims were: (1) evaluate the potential of Watson's framework for studying species composition in fragments and islands for a specific landscape type: cryptobiotic crust systems in the arid south-western US; (2) expand Watson's original model to include ephemeral/non-equilibrium systems by revising his categories of patch age and matrix contrast; and (3) examine the interplay between patch dynamics and species autecology, demonstrating the need for more work on ephemeral patchy systems. Location Cryptobiotic crust systems in two piñon-juniper sites in Central New Mexico, western North America. Results Watson's patch age designation was not applicable to our system because of its ephemeral or non-equilibrial nature. Based on this result, we constructed what we refer to as a ‘speed key’ that includes equilibrium and non-equilibrium patches of all kinds. For this model we maintained one of Watson's original traits: patch origin, and amended two others to describe persistence and permeability across the matrix. Importantly, persistence and matrix permeability must be evaluated as functions of the organisms under consideration. Systems that may be in equilibrium for one taxon may well be non-equilibrial (ephemeral) for another. A patch that appears to be in high contrast with its intervening matrix may actually be in low contrast, depending on the dispersal ability of the organism through that matrix. Main conclusions To improve substantially on our understanding of patchy systems (whether islands or fragments) it is important to account explicitly for relevant organismal life-history traits in the designation of those systems. Too often, patches are defined by how the researcher views them from his/her own spatio-temporal viewpoint. Once we move to an organism-centred understanding of these patches we may find surprising and novel comparisons that allow us to move across scales and inform our view of ecological patterns and processes. By incorporating non-equilibrium systems into a model of insularity, this work has general implications that go beyond the scope of cryptobiotic crusts to add to the current dialogue in biogeography.