It's now almost six years since we followed through on a madcap plan to warm up an entire stream reach at Hengill. It was only made possible by the ingenuity of our engineer, Philip Johnson, shown here on the right (Alex Huryn is providing the muscle behind him). Philip is now retired, but still working with us (and still just for the kicks too). Happily, all that work was worth it. This month Dan Nelson has published two papers from the project. The new one, published in Ecology, extends the results of his recent paper in Global Change Biology to show that secondary production in the manipulated stream remained constant after warming, as predicted by metabolic theory. The mechanism was surprising, however. Rather than total biomass declining and mean growth rates increasing, total biomass did not change. Instead small fast-growing taxa declined as larger slow-growing species, including newly invading taxa, increased in biomass. As shown in his previous paper, Dan's results were driven by somewhat idiosyncratic relationships between body size and thermal preference, and demonstrate how reassembly of communities under warming will depend on the range in thermal preferences and dispersal abilities in regional species pools, which may or may not be related to body size. This is the kind of result that could only be observed with an experimental manipulation of a natural ecosystem that is open to dispersal processes. Which is one reason why we go to all the trouble!