MacArthur and Wilson II: Archipelagos

Last weekend, my post outlined MacArthur and Wilson's Theory of Island Biogeography for one island. Recall that we discussed on how islands gained species, through colonization, and how they lost species, through extinction.

But islands often occur as part of an archipelago, a collection of islands. Islands in the archipelago vary in a number of ways, including topography, shape, size, and distance to the mainland. MacArthur and Wilson focused on the effect that two characteristics affect diversity: distance between the island and the mainland, and island size.

Click to enlarge

The distance between the island and the mainland would mainly affect colonization. Islands that are close to the mainland are relatively easy targets -- in fact, an animal might even be able to see them. Even random departures from the mainland favor colonization of a nearby island over a far-away island because a wider range of departure directions encompass the island if it is close-by than if it is far away. So a seed blown off the mainland or a bird blown out to sea would be more likely to land on a near island than a far island. The journey to a close island would be easier as well -- a displaced animal or seed would be more likely to survive a short journey to an island than a long one. Distance, however, should affect extinction very little.

Island size would affect extinction rate but have a relatively small effect on colonization rates.  Larger islands could support larger populations, which are less likely to go extinct just by chance than smaller populations. Larger islands also should have greater habitat diversity, offering many environments for organisms to inhabit. New colonists, then, would be more likely to find the kind of habitat they need to survive. So extinction rates should be lower on large islands and higher on smaller islands.

We can modify our graphical version of the model to accommodate these new wrinkles. We now have two lines for colonization -- a higher colonization rate for islands that are close to the mainland and a lower rate for islands that are far from the mainland. Likewise, we have two lines for extinction: a lower rate for large islands, a higher rate for small islands.

MacArthur and Wilson Theory of Island Biogoegraphy

Now there are four crossing points and four equilibria.  Overall, MacArthur and Wilson's model predicts:

• Small islands far from the mainland have the lowest biodiversity at equilibrium (equilibrium point A); 
• Large islands close to the mainland have the highest biodiversity at equilibrium (equilibrium point D); and
• The other two combinations (small islands close to the mainland, large islands far from the mainland) would have intermediate biodiversity at equilibrium (equilibrium points B and C).

I've drawn the colonization and extinction curves so that they look about the same, but that was only for convenience. The shape of the two curves can look quite different from what I've drawn and from each other. This will affect the predicted species diversity at equilibrium, but not the overall conclusion: large, near islands have higher diversity at equilibrium than small, far islands.

How does diversity vary across islands of different sizes and distances from the mainland?  The data are somewhat mixed, but overall increasing size and increasing distance are correlated with higher and lower diversity, respectively. For example, Brown and Peck (1996) studied the diversity of cerambycid beetles on islands of the Florida Keys. They found that diversity generally increases with increasing island size and decreases as distance to the mainland increases. 

Species diversity and island size (left) and distance from the mainland (right)
for cerambycid beetles in the Florida Keys.
Click to enlarge
(See image credits for attribution)

It is harder to investigate the interplay between size and distance.  In the California Channel Islands, for example, the nearer islands also tend to be larger, confounding the two factors.

This model was built for oceanic islands -- those that have never been attached to a continent. What about land-bridge islands? They do not start out empty of organisms, they start out full!

Ideally, when a land-bridge island first separates from the mainland, it has the full complement of species found in that area. But over time, the island undergoes relaxation, a process by which species are lost. Of course, new species also arrive from the mainland and these would supplement diversity on the land-bridge island. The predicted biodiversity at equilibrium would conform to the MacArthur and Wilson model.

The Panama Canal
Click to enlarge

The trend over time for land-bridge islands, however, is very different than for oceanic islands. Instead of gaining biodiversity over time, these kinds of islands lose biodiversity. A great example of diversity relaxation is Barro Colorado Island (BCI) in Panama. It was formed in 1923 when waters from the newly-dammed Rio Chagres created Lake Gatun, an integral part of the Panama Canal. The rising waters converted a rainforest hilltop into a rainforest island. Since the island was formed,  BCI has lost biodiversity, including as many as 60 species of birds.

References:

Brown, D. J. and S. B. Peck. 1996. The biogeography of the Cerambycidae (Coleoptera) of the Florida Keys, the Bahama Islands and Cuba. Canadian Journal of Zoology 74: 2154-2169.

Karr, J. R. 1982. Avian extinction on Barro Colorado Island, Panama: A reassessment. The American Naturalist. 119: 220-239.

MacArthur R. M. and E. O. Wilson. 1967. The Theory of Island Biogeography. Princeton University Press: Princeton, NJ.

Image credits:
Panama Canal map:  http://ctfs.arnarb.harvard.edu/webatlas/datasets/bci/

Cerambycid beetle figures adapted from: http://www.tiem.utk.edu/~gross/bioed/bealsmodules/spec_area.html