QUEST, University of Bristol, UK

A theoretical biologist suggests that evolution makes plants more predictable.

The debate over how forests respond to rising levels of carbon dioxide has brought home to me how much spin even a dry journal article can contain.

In the mid-1990s, when the forest Free Air Carbon dioxide Enrichment (FACE) experiments began, I thought that we were poised to learn how trees really respond to carbon dioxide. In these experiments, carbon dioxide is pumped over forests to simulate future conditions.

Unfortunately, years of data collection and scores of papers later, we still haven't reached agreement. Using the same data, researchers conclude that carbon dioxide either fertilizes forests or it doesn't (or the effect is small, or it goes away, or will soon go away...)

The situation would be helped if we had better theories of how trees might be expected to react to changes in their resources. It was refreshing, therefore, to encounter an elegant analysis of plant behaviour (O. Franklin New Phytol. doi:10.1111/j.1469-8137.2007.02063.x; 2007).

Plants, subject to selective pressure, have to optimize what they can. This is a basic principle of evolutionary biology, too often disregarded in experimental contexts.

Theoreticians have long known that an individual leaf in high carbon dioxide will maximize the amount of carbon it fixes — a measure of its growth success — if it lowers its nitrogen content to optimize the balance between photosynthesis and respiration.

Franklin extends this nitrogen optimization principle to the whole plant, a significantly more complex problem. His model predicts 83% of the variation in plant growth enhancement seen across FACE studies, explains the observed relationship between plant growth and canopy nitrogen content, and does much else besides. It is a welcome step forwards.