Marshall chose to study Hawaii's biological history to test the theory that declines in landmass can have more severe impacts on biological diversity than previously thought. The discovery that some mammal groups had declined over periods of millions of years puzzled biologists, since diversity is expected to increase in the absence of catastrophic events, and overcrowding was raised as a possible explanation.
Although Hawaii has no native mammals, the islands' diverse ages and shrinking sizes – Maui Nui is now two-thirds smaller than it was at its peak – provided a good opportunity to test the idea across 14 other groups of plants and animals.
The challenge was to estimate the past diversity since Hawaii is largely bereft of fossils. Marshall used models of how the species' richness would grow and decline as the islands' area waxed and waned, and compared them with what we know of the current richness of each island using the differences in island ages to see which models described reality best.
The implications may extend beyond islands. The paper refers to moist forests and the Eastern Pacific marine tropics as ecosystems that have contracted over millions of years, and may have lost diversity in the process.
I'iwi, or scarlet honeycreepers, have survived despite the diversity of their genus declining as the Hawaiian islands have shrunk, contradicting the expectation that biological diversity grows with time. Nate Yuen