In a rather elegant piece from our recent environmental issue, Jonathan Rosen asked about endangered species: "How do we know what to save, at what cost, and why?" It's a good question, and Rosen smartly notes that many of our conservation efforts are biased toward the cute and cuddly. In a related vein, Tyler Cowen quotes Jeff Sachs on the importance of preserving biodiversity (which, granted, isn't always the same thing as species extinction):

The main lesson of ecology is the interconnectedness of the various parts of an ecosystem and the dangers of abrupt, nonlinear, and even catastrophic changes caused by modest forcings... It is a basic finding that biological diversity increases the productivity and resilience of ecosystems. With more species filling more niches in a given location, a biodiverse ecosystem is better buffered against external shocks in is more adept at cycling nutrients, capturing solar radiation, utilizing water resources, and preventing the takeover of the system by single predators, weeds, or pathogens.

In other words, preserving biodiversity helps to preserve all aspects of ecosystem functions. Removing one or more species from an ecosystem, for example, by selective harvesting of trees or fish or hunted animals, can lead to a cascade of ecological changes with large, adverse, and nonlinear effects on the functioning of the ecosystem.

Cowen says he'd agree with those claims, but wonders if this could be phrased in economic terms: "I still cannot articulate to a skeptic exactly what marginal disaster will come if we do not take drastic action to preserve biodiversity... What are these costs as a percentage of GDP?" His commenters have an array of thoughtful responses. Frankly, I'm stumped as to how best to think about this question—or whether it's even the right question—but here's a short, readable paper from the World Bank that gets into the difficulty of trying to value ecosystems in economic terms. (For starters: "Economic valuation tends to handle very large-scale and long-term problems rather poorly.") Good discussion, though.

--Bradford Plumer