MacArthur and Wilson suggested that the biodiversity of an island will vary in direct proportion to a function of the island's size (i.e., larger islands can support a greater number of species) and in inverse proportion to a function of its distance from the mainland (i.e., many remote islands will tend to support fewer species). Reduced biodiversity in an island context is likely to require significant adaptation on the part of colonizing human populations. Evans argues that this limitation makes islands ideal laboratories for the study of human adaptations to the natural environment, whilst Renfrew and Wagstaff, in the introduction to their study of Melos, focus on this limitation in biodiversity as a "significant characteristic of the island ecosystem." For human communities, however, this limitation may potentially be offset by other factors. The reduced biodiversity of an island ecosystem applies only to terrestrial resources: the resources of the sea will be as rich as on any other coastal area, and may be equally important to human communities. A small island such as Malta or Melos allows all communities direct access to the sea, providing an important nutritional "safety net," as well as an element of dietary diversity, which may actually give island communities an advantage over their landlocked counterparts. Islands may also have specific non-biological resources (such as obsidian on Melos), which may be used in exchange with communities on other islands and adjacent mainlands.