For over 150 years, insular environments have played a key role in biology as they provide an explicit spatial and temporal context in which to study the processes behind biodiversity. Here I propose to use marine lakes - islands of seawater surrounded by land - as sample habitats to address the role of isolation in marine biodiversity. Marine lakes share many analogies with island systems: they are well-defined geographically, harbor unique biota with high numbers of endemic species and rapidly evolving populations. By comparing six marine lakes in Indonesia with comparable ages and sizes, but differing in degrees of connection to the adjacent sea I aim to establish whether short-term isolation leads to an increase in local species and genetic diversity. I will apply a multi-level comparative approach based on 1) spatial patterns of present day lake-species assemblages, 2) historic patterns of species diversity since formation of the lakes using sediment cores, 3) genetic patterns of diversity. This unique combination of study system, data and methodology will allow me to establish the role of isolation in dictating the relative contribution of ecological versus evolutionary processes on marine lake community assembly.