Photo: Tropical group of islands in Ang Thong National Marine Park. GVS / Adobe Stock

Photo: Tropical group of islands in Ang Thong National Marine Park. GVS / Adobe Stock

A new paper in Proceedings of the National Academy of Sciences co-authored by Professor Rob Whittaker provides the first analysis to test the combined effects of underlying processes on the emergent patterns of island diversity and reveals that islands' biodiversity is less affected by their remoteness than expected.

Identifying global drivers of species diversity has long been a central goal in geographical ecology and biogeography. One long established idea is that with increasing isolation of discrete areas like islands from similar habitats, their species richness decreases. This effect is thought to be strongest when looking at very small and remote islands, where the low environmental complexity and the lack of well-developed, complex food webs leads to relatively higher levels of local population failure and extinction counterbalanced by comparatively low levels of recruitment of new arrivals. An otherwise similar small island that is near to a continent, with its vast pool of potential colonists, will, in contrast, receive so many chance arrivals of new species that their richness at any one time, although less than nearby large islands, will be comparatively high.

The upshot is that the relationship between island area and species richness (which is one of the few accepted general ecological laws), should vary systematically with the isolation of the archipelago; with steeper slopes of increase and from lower start points, across a set of small to large islands as the distance from the mainland source pool increases (below). Variation in the island species–area relationship (the ISAR), is not merely of academic interest as the ISAR also provides a key tool in predicting species losses as land use change reduces and fragments areas of pristine habitats within continents and islands alike. The question is, how well does this simple theory work?

Figure: Illustration of how the island species–area relationship is expected to change with increasing isolation (down-pointing arrow) from a mainland source pool.

Figure: Illustration of how the island species–area relationship is expected to change with increasing isolation (down-pointing arrow) from a mainland source pool. This effect is hard to detect in comparative analyses of a large number of separate archipelago datasets, in large part because the dynamics of colonization (and speciation) and extinction are responsive to the disposition of area more locally, within the archipelago and often between archipelagos.

In their paper just published in the journal Proceedings of the National Academy of Sciences of the USA, an international team of biogeographers, Thomas Matthews, Francois Rigal, Kostas Triantis and Robert Whittaker, show that the expected effect of isolation from mainland source areas is typically swamped by other factors. The authors used a form of path analysis, a statistical approach that allows the interactions between different factors to be explored, and they applied it to a large set of 151 separate ISARs, each describing the increase in species richness with increasing island area for a particular type of organism (plants, birds, mammals, land snails, spiders, etc) on a particular archipelago.

Co-author Robert Whittaker, Professor of Biogeography at the School of Geography and the Environment, Oxford, comments “Some of the variation in the final models is just the result of differences between different types of organism, with richness of plants unsurprisingly greater than lizards or mammals on islands. More interestingly, it seems that the internal configuration of the archipelago itself is more important than how far it is from a mainland source pool in determining how many species each island holds.” Here again, the area of the archipelago itself is critical, but also how that total area is distributed. The model seems to show that the overall species diversity of an archipelago for a given group of organisms is very largely a function of the combined area of each island. Lead author Thomas Matthews (University of Birmingham and Honorary Research Associate at the School of Geography and the Environment) adds, “The results tell us that for a given size of archipelago, the form of the island species–area is quite variable, but that this variation describes a trade-off between increasing steepness of slope and diminishing richness of the smallest islands, which reflect how much the total land area of the archipelago is subdivided and scattered across the islands that make up that archipelago.”

Kostas Triantis, also an author on the paper and a biologist from the National & Kapodistrian University of Athens, Greece adds “…in addition to revealing the importance of within-archipelago processes to how many species each island holds, the lack of a mainland isolation effect in our models tells us something else about the processes building island biodiversity, which is about the contribution of evolution on islands. Ever since Darwin we have known that remote islands hold more endemic species: plants and animals that have evolved in situ. While we can actually see slightly steeper island species-area relationships if we separate out the oceanic islands, such as Hawaii and Galápagos, from the rest, the effect is not as strong as originally expected. This is because, the origination of new species on remote islands is, in time, strong enough to increase the diversity of the communities on those islands, largely compensating for the slow pace of immigration.” While the implications of evolution on islands have long been understood, this paper provides the first analysis to test the combined effects of all these underlying processes on the emergent patterns of island diversity.

The authors conclude that their findings point to a need for further research to quantify how different scales of isolation (within and between archipelagos, and among meta-archipelagos) work to determine the exchange of species across insular systems, in order to develop improved models that can predict how species diversity changes as habitats become more isolated and fragmented, whether naturally or through human action.