The Voltage Gate

This seems to be a more sensible theory regarding leaf color change in autumn:

By taking careful stock and laboratory analyses of the autumn foliage of sweetgum and red maple trees along transects from floodplains to ridge-tops in a nature preserve in Charlotte, N.C., former University of North Carolina at Charlotte graduate student Emily M. Habinck found that in places where the soil was relatively low in nitrogen and other essential elements, trees produced more red pigments known as anthocyanins.

Habinck's discovery supports a 2003 hypothesis put forward to explain why trees bother to make red pigments, by plant physiologist William Hoch of Montana State University, Bozeman. Hoch found that if he genetically blocked anthocyanin production in red-leafed plants, their leaves were unusually vulnerable to fall sunlight, and so sent less nutrients to the plant roots for winter storage.

For trees living in nutrient-poor soils, then, it makes sense to produce more anthocyanins, which protect the leaves longer, so as much nutrient as possible can be recovered from leaves before winter sets in. It is, after all, the process of recovering of nutrients from leaves which turns leaves from green to yellow, orange and sometimes anthocyanin-red.

The trees Habinck studied appear to be acting in accordance with Hoch's hypothesis. "It makes sense that anthocyanin production would have a function, because it requires energy expenditure," said Habinck. Put in economic terms, anthocyanins are an investment made by stressed trees in situations where they stand to gain from the extra recovery of nutrients from leaves. It's not about the showy color, but about survival.