Wikiecology

When I saw the graph on the front of today's technology Guardian, in an article about the slowing growth of Wikipedia (it's not in the online version), I thought: "That looks like a logistic growth curve. Perhaps the sum total of knowledge represents a resource that is being exhausted, causing the encyclopedia's growth to slow".

When I got to the end of the article, I discovered that the researcher behind the work being discussed, Ed Chi of PARC, had a similar thought.

"In my experience, the only thing we've seen these growth patterns [in] before is in population growth studies – where there's some sort of resource constraint that results in this model." The site, he suggests, is becoming like a community where resources have started to run out. "As you run out of food, people start competing for that food, and that results in a slowdown in population growth and means that the stronger, more well-adapted part of the population starts to have more power."

But the article also says that the slowdown is caused by a shift in power towards 'deletionist' editors - it's getting harder to get stuff onto Wikipedia. Which suggests that the correct model might not be a sort of density-dependent, resource-limited population (we're running out of stuff to create wikipedia entries about), but perhaps something more top-down, like a predator-prey system (the editor population is keeping the contributor population in check).

I've no idea, but it's a question entirely suitable to the tools of ecological analysis. Someone should get onto it.

Just Science day 5: Kill and cure?

The best way to protect an endangered species might be to ... cull it.

That's the conclusion reached by Michael Bode and Hugh Possingham in a theoretical analysis published recently in Ecological Modelling. The problem that they address is the tendency of some species to show boom-and-bust population dynamics. This seems a particular feature of tightly linked predator-prey systems: if predator is rare, prey booms, predator has lots of food, predator booms, prey crashes, predator starves, predator crashes, and so on. The Canadian lynx-hare system is the most famous example.

At the bottom of the cycles, population sizes might be so small that either predator or prey is highly vulnerable to extinction - by some other event, such as disease or disaster.

Bode and Possingham model various options for damping these cycles by removing some predators, prey, or both, from the system. "Remarkably," they say, "if the interventions are enacted at the appropriate time, infrequent culling of a small number of individuals significantly reduces the probability of extinction of the predator."

My first thought when I saw this, was that we probably rarely know enough about a species' population dynamics to know when the best time to cull is, and that such a move would be politically untenable. But, as Bode points out, the most likely candidates for such a move are large game-park species, such as elephants, about which we know quite a lot. It's also heartening that their (as they say, simplistic) simulation suggests that culling prey (which would be plants for elephants, I suppose) is more likely to be effective than predators, and that culling would be needed only rarely.

They conclude:

Most current methods of ecosystem control attempt to keep a single species population below a certain threshold, or between fixed bounds, leading to highly inefficient management. Furthermore, unnecessarily high levels of culling can generate considerable public opposition. Given the clumsy nature of such static controls, this view may be reasonable. Our dynamic control solution has the substantial benefit of minimising the preventable culling of organisms, and being extremely cost-effective.

I swapped e-mails with Bode, and besides mentioning that, in parks with very high elephant densities, such as Kruger, culling elephants seems 'almost unavoidable' (the issue seems to have flared up most recently in late 2005, judging by this BBC story), he also mentioned that similar issues were arising in Australia regarding koalas and kangaroos. As he says, why is it always the cure animals?

All I can say is I'm glad I don't have to make these decisions. Culling animals is clearly a terrible business - particularly for long-lived, highly social species such as elephants. But we now effectively control the destiny of all large, slow-breeding vertebrates. If the only places they can survive is in the parks we set up, and if we take a truly utilitarian and, dread word, holistic view of conservation, then we also have to take the responsibility of 'managing' their populations. Science isn't he only thing to guide us on such decisions, but it's best ot take as informed and clever a decision as possible, towards which Bode and Possingham's ideas can only help.